Technical Report Documentation Page

1. Report No.

2. Government Accession No.

FWHA/TX-03/1774-2
4. Title and Subtitle

3. Recipients Catalog No.

5. Report Date

October 2001

Effects of Wrapping Chloride Contaminated

Concrete with Fiber Reinforced Plastics

6. Performing Organization Code

7. Author(s)

8. Performing Organization Report No.

Research Report 1774-2

E. W. Berver, J. O. Jirsa, D. W. Fowler, H. G. Wheat, and T. Moon

9. Performing Organization Name and Address

10. Work Unit No. (TRAIS)

11. Contract or Grant No.

Center for Transportation Research

The University of Texas at Austin
3208 Red River, Suite 200
Austin, TX 78705-2650

Research Study 0-1774

12. Sponsoring Agency Name and Address

13. Type of Report and Period Covered

Texas Department of Transportation

Research and Technology Transfer Office
P.O. Box 5080
Austin, TX 78763-5080

Research Report (9/97-1/00)

14. Sponsoring Agency Code

15. Supplementary Notes

Project conducted in cooperation with the U.S. Department of Transportation, Federal Highway Administration, and
the Texas Department of Transportation.
16. Abstract

Damage to concrete due to corrosion of steel reinforcement is a costly maintenance problem that affects
infrastructure. Reinforced concrete structures located in an aggressive environment are susceptible. Fiber reinforced
plastic composite wraps have recently been used to rehabilitate structures that have experienced damage due to
corrosion. Little is known about the long-term performance of FRP composites in corrosion prevention. Corrosion
monitoring of laboratory specimens and field research are discussed. A new phase of the project involving the use
of corrosion inhibitors prior to wrapping with FRP composites is also described.
17. Key Words

18. Distribution Statement

corrosion, steel, concrete, FRP, polymer wrap,

wrapping, fiber-reinforced plastics, fiberreinforced polymers, repair, rehabilitation,
encapsulation
19. Security Classif. (of report)

Unclassified

No restrictions. This document is available to the public through

the National Technical Information Service, Springfield, Virginia
22161.

20. Security Classif. (of this page)

Unclassified
Form DOT F 1700.7 (8-72)

21. No. of pages

22. Price

112
Reproduction of completed page authorized

EFFECTS OF WRAPPING CHLORIDE CONTAMINATED

CONCRETE WITH FIBER REINFORCED PLASTICS

by

E. W. Berver, J. O. Jirsa, D. W. Fowler,

H. G. Wheat, and T. Moon

Research Report No. 1774-2

Research Project 0-1774
Effect of Wrapping Chloride Contaminated Structural Concrete
with Multiple Layers of Glass Fiber/Composites and Resin

conducted for the

Texas Department of Transportation

in cooperation with the

U.S. Department of Transportation
Federal Highway Administration
by the
CENTER FOR TRANSPORTATION RESEARCH
BUREAU OF ENGINEERING RESEARCH
THE UNIVERSITY OF TEXAS AT AUSTIN

October 2001

Research performed in cooperation with the Texas Department of Transportation and the U.S.
Department of Transportation, Federal Highway Administration.

ACKNOWLEDGMENTS
We greatly appreciate the financial support from the Texas Department of Transportation that made this project
possible. The support of the Project Director, Jon Kilgore, (SAT) and Project Coordinator, W. Cox, (BRG) is
also very much appreciated. We would also like to thank monitoring committee members, R. Baker (LBB),
David McDonnold (BRG), and Keith Ramsey (BRG) for their guidance on this project.

DISCLAIMER
The contents of this report reflect the views of the authors, who are responsible for the facts and the accuracy of
the data presented herein. The contents do not necessarily reflect the view of the Federal Highway
Administration or the Texas Department of Transportation. This report does not constitute a standard,
specification, or regulation.

NOTICE
The United States Government and the state of Texas do not endorse products or manufacturers. Trade or
manufacturers names appear herein solely because they are considered essential to the object of this report.

NOT INTENDED FOR CONSTRUCTION,

BIDDING, OR PERMIT PURPOSES
James O. Jirsa, Texas P.E. No. 31360
David Fowler, Texas P.E. No. 27859
Harovel Wheat, Texas P.E. No. 78364
Tess Moon
Research Supervisors

iv

TABLE OF CONTENTS
CHAPTER 1: INTRODUCTION ............................................................................................................... 1
1.1 INTRODUCTION TO CORROSION IN REINFORCED CONCRETE BRIDGES .................................. 1
1.2

PREVIOUS RESEARCH ............................................................................................................. 2

1.3

PROJECT OBJECTIVES ............................................................................................................. 3

CHAPTER 2: CORROSION OF STEEL IN CONCRETE.......................................................................... 5

2.1 MECHANISMS OF CORROSION ................................................................................................ 5
2.2

PREVENTION OF CORROSION .................................................................................................. 7

2.3

REPAIR METHODS ................................................................................................................... 8

CHAPTER 3: FIBER REINFORCED PLASTIC COMPOSITES ............................................................... 11

3.1 INTRODUCTION TO FRP COMPOSITES .................................................................................. 11
3.2

SEISMIC APPLICATIONS ........................................................................................................ 13

3.3

APPLICATIONS IN CORROSION REPAIR AND PREVENTION ................................................... 14

3.4

LONG-TERM BEHAVIOR........................................................................................................ 16

CHAPTER 4: CORROSION MONITORING METHODS ......................................................................... 19

4.1 INTRODUCTION TO CORROSION MONITORING...................................................................... 19
4.2

VISUAL INSPECTION ............................................................................................................. 19

4.3

HALF-CELL POTENTIAL ......................................................................................................... 19

4.4

LINEAR POLARIZATION ......................................................................................................... 21

4.5

LONG-TERM MONITORING ................................................................................................... 25

CHAPTER 5: PROJECT ORGANIZATION ............................................................................................ 27

5.1 TXDOT PROJECT 1774......................................................................................................... 27
5.2

5.3

LABORATORY PHASE ............................................................................................................ 27

5.2.1 Specimen Geometry.................................................................................................. 27
5.2.2 Concrete.................................................................................................................... 29
5.2.3 Simulation of Existing Field Conditions .................................................................. 30
5.2.4 Wrapping Materials and Procedures ........................................................................ 31
FIELD PHASE ......................................................................................................................... 33

CHAPTER 6: LABORATORY DATA ..................................................................................................... 39

6.1 SCOPE OF PROJECT ............................................................................................................... 39
6.2

6.3

CORROSION MONITORING .................................................................................................... 39

6.2.1 Half-Cell Potential Results ....................................................................................... 41
6.2.2 Linear Polarization Results ...................................................................................... 44
TESTING OF SPECIMENS ........................................................................................................ 47
6.3.1 Crack Mapping ......................................................................................................... 47

6.4

6.3.2 Chloride Determination ............................................................................................ 48

6.3.3 Bond Test.................................................................................................................. 50
OPENING OF SPECIMENS, GROUP A...................................................................................... 51

6.5

OPENING OF SPECIMENS, GROUP B...................................................................................... 53

6.6

DISCUSSION OF RESULTS ...................................................................................................... 56

CHAPTER 7: FIELD DATA .................................................................................................................. 59

7.1 SCOPE OF PROJECT ............................................................................................................... 59
7.2

FIELD RESULTS PRIOR TO REPAIRS ...................................................................................... 59

7.3

MONITORING RESULTS ......................................................................................................... 61

7.4

OBSERVATION OF FIELD INSTALLATION............................................................................... 63

7.5

COMPARISON OF RESULTS FROM THE LABORATORY AND THE FIELD ................................. 66

CHAPTER 8: CORROSION INHIBITORS .............................................................................................. 69

8.1 INTRODUCTION TO CORROSION INHIBITORS......................................................................... 69
8.2

TYPES OF CORROSION INHIBITORS ....................................................................................... 69

8.3

DESCRIPTION OF TEST SPECIMENS WITH CORROSION INHIBITORS ...................................... 70

8.4

MONITORING AND EXPOSURE .............................................................................................. 73

CHAPTER 9: SUMMARY AND CONCLUSIONS .................................................................................... 79

9.1 SUMMARY ............................................................................................................................. 79
9.2

9.3

9.4

REVIEW OF FINDINGS ........................................................................................................... 79

9.2.1 Corrosion Activity .................................................................................................... 79
9.2.2 Role of FRP as a Barrier........................................................................................... 79
9.2.3 Effect of Repair Materials ........................................................................................ 79
9.2.4 Field Observations.................................................................................................... 80
PRELIMINARY CONCLUSIONS AND RECOMMENDATIONS ..................................................... 80
9.3.1 Effectiveness of FRP in Corrosion Prevention......................................................... 80
9.3.2 FRP as a Barrier........................................................................................................ 80
9.3.3 Corrosion Monitoring............................................................................................... 80
9.3.4 Recommendations for Field Applications ................................................................ 80
RECOMMENDATIONS FOR CONTINUING RESEARCH ............................................................. 81

APPENDIX A: DETAILS OF TEST SPECIMENS .......................................................................... 83

APPENDIX B: HALF-CELL POTENTIAL READINGS 2/99-4/00 .......................................................... 87
APPENDIX C: PHOTOGRAPHS OF STEEL REINFORCING FOR GROUP A ....................... 89
REFERENCES........................................................................................................................................ 97

vi

LIST OF FIGURES
Figure 1.1

Corrosion damage to a bridge bent ................................................................................... 2

Figure 1.2

Bridge bent and column wrapped with FRP after repair of corrosion damage................. 2

Figure 2.1

The corrosion process (Broomfield 1997) ........................................................................ 5

Figure 2.2

Consequences of corrosion of steel in concrete................................................................ 6

Figure 3.1

FRP composite wrap being applied to the bent of a bridge ............................................ 12

Figure 3.2

Highway bridge column after the 1994 Northridge earthquake (Courtesy of EERC,
University of California, Berkeley). ............................................................................... 14

Figure 3.3

Concrete pile that had previously been repaired with a fiberglass jacket....................... 16

Figure 4.1

Copper-Copper Sulfate Half-cell Circuitry (ASTM 1989) ............................................. 20

Figure 4.2

Equipotential contour diagram (ASTM 1989)................................................................ 21

Figure 4.3

Hypothetical anodic and cathodic polarization curve..................................................... 22

Figure 4.4

Schematic of linear polarization device (Broomfield 1997)........................................... 23

Figure 4.5

Schematic of linear polarization device with a sensor controlled guard ring

(Broomfield 1997)........................................................................................................... 24

Figure 5.1

Geometry of cylinder specimens..................................................................................... 28

Figure 5.2

Typical cylinder specimen .............................................................................................. 28

Figure 5.3

Geometry of rectangular block specimens...................................................................... 29

Figure 5.4

Typical rectangular block specimen ............................................................................... 29

Figure 5.5

Cracked cylinder specimen ............................................................................................. 30

Figure 5.6

Wrapped cylinder specimen............................................................................................ 32

Figure 5.7

Wrapped rectangular block specimen ............................................................................. 33

Figure 5.8

Damage to an endcap of Structure #5 ............................................................................. 34

Figure 5.9

Delamination on an endcap of Structure #8.................................................................... 34

Figure 5.10 Cracking on column of Structure #12 ............................................................................. 35

Figure 5.11 Structure #1 prepared for encapsulation ......................................................................... 36
Figure 5.12 Structure #2 encapsulated with FRP wrap ...................................................................... 36
Figure 5.13 Concorr corrosion rate probe and connection cable (Concorr 1998).............................. 37
Figure 5.14 Lengthwise section of the corrosion rate probe (Concorr 1998) .................................... 37
Figure 5.15 Installed probe on Structure #7 ....................................................................................... 38
Figure 6.1

Cylinder specimens in the exposure tank........................................................................ 39

Figure 6.2

Longitudinal cracks forming on an uncracked specimen................................................ 41

Figure 6.3

Half-cell potential vs. time for Group A ......................................................................... 41

vii

Figure 6.4

Half-cell potential vs. time for Group B ......................................................................... 42

Figure 6.5

Uncontaminated concrete cylinders with different parameters ...................................... 42

Figure 6.6

Chloride contaminated concrete cylinders with different parameters ............................ 43

Figure 6.7

Half-cell potential vs. time for rectangular blocks.......................................................... 43

Figure 6.8

PR-Monitor test setup on specimen CC5........................................................................ 45

Figure 6.9

Corrosion current density (A/cm2) values for Group B................................................ 46

Figure 6.10 Evidence of corrosion activity in the splash zone of specimen CNC10 ......................... 47
Figure 6.11 Test setup for DYNA Z16 Pull-Off Tester ..................................................................... 50
Figure 6.12 Failure at the epoxy/concrete interface on Test #2 ......................................................... 51
Figure 6.13 Moisture trapped beneath the surface in the splash zone of specimen CNC14.............. 53
Figure 6.14 Section loss on the upstream stirrup of specimen RNC6................................................ 55
Figure 6.15 Macrocell formation in the top portion of specimen CC5 .............................................. 55
Figure 6.16 Steel reinforcing of CNC10 with a corrosion rate of 3.86 mpy...................................... 57
Figure 6.17 Steel reinforcing of CC6 with a corrosion rate of 1.41 mpy........................................... 57
Figure 6.18 Steel reinforcing of CC3 with a corrosion rate of 0.34 mpy........................................... 58
Figure 7.1

Corrosion current density (A/cm2) values from different equipment ........................... 60

Figure 7.2

PR-Monitor test setup for Structure #12......................................................................... 61

Figure 7.3

Corrosion current density (A/cm2) for corrosion monitoring ....................................... 63

Figure 7.4

Finish of FRP wrap around girders of Structure #12...................................................... 64

Figure 7.5

Finish of FRP wrap around girders of Structure #7........................................................ 64

Figure 7.6 Finish of FRP wrap around girders of Structure #8.......................................................... 65

Figure 7.7 Vehicular collision damage on column of Structure #2 ................................................... 65
Figure 7.8

Rust stains on bent of Structure #1 ................................................................................. 66

Figure 8.1

Application of a surface applied corrosion inhibitor to the specimens .......................... 71

Figure 8.2

Filling voids with a thickened epoxy filler ..................................................................... 72

Figure 8.3

Saturating glass fabric with epoxy resin ......................................................................... 72

Figure 8.4

Applying FRP wrap to the concrete specimen................................................................ 73

Figure 8.5

Chloride determination of VETEK System .................................................................... 74

Figure 8.6

Installation of the embedded electrodes.......................................................................... 74

Figure 8.7

Wires from the embedded probes and the connection to the steel.................................. 75

Figure 8.8

Specimens in the exposure tank. ..................................................................................... 77

viii

LIST OF TABLES
Table 2.1

Recommended limits for water-soluble chloride ion content in concrete, from ACI
201.2R-77 ........................................................................................................................... 7

Table 3.1

Basic Properties of Glass, Carbon, and Aramid FRP Compared to Steel (Bassett
1998) ................................................................................................................................. 11

Table 4.1

ASTM interpretation of half-cell readings ....................................................................... 21

Table 4.2

Interpretation of Linear Polarization Results ................................................................... 24

Table 4.3

Correlation of corrosion rates to rust growth and section loss ......................................... 24

Table 5.1

Concrete mix design ......................................................................................................... 30

Table 5.2

Location of structures repaired with FRP composite ....................................................... 33

Table 5.3

Probe installation locations............................................................................................... 38

Table 6.1

Specimen Parameters........................................................................................................ 40

Table 6.2

Comparison of Half-Cell Potential and PR-Monitor ........................................................ 44

Table 6.3

Linear Polarization Data from Group B ........................................................................... 46

Table 6.4

Crack Size Data ................................................................................................................ 48

Table 6.5

Results from chloride tests for Group A at 1 -in. depth ................................................ 49

Table 6.6

Results from chloride tests for Group B........................................................................... 49

Table 6.7

Results from pull-off test for RNC6 ................................................................................. 51

Table 6.8

Observations from opening of Group A ........................................................................... 52

Table 6.9

Observations from opening of Group B ........................................................................... 54

Table 7.1

Corrosion Rate Measurement Data (July 1998) ............................................................... 59

Table 7.2

Chloride percentages by weight of concrete (FSEL 1999)............................................... 60

Table 7.3

Linear Polarization Data................................................................................................... 62

Table 8.1

Specimen Parameters for Corrosion Inhibitor Study........................................................ 70

Table 8.2

Interpretation of Readings for the VETEK System.......................................................... 73

Table 8.3

Half-cell Potential Readings ............................................................................................. 76

Table 8.4

Embedded Probe Readings ............................................................................................... 76

Table 8.5

Chloride Measurements from Embedded Probes ............................................................. 77

ix

SUMMARY
Damage to concrete due to corrosion of steel reinforcement is a costly maintenance problem that
affects infrastructure. Reinforced concrete structures located in an aggressive environment are
susceptible. Fiber reinforced plastic composite wraps have recently been used to rehabilitate
structures that have experienced damage due to corrosion. Little is known about the long-term
performance of FRP composites in corrosion prevention. Corrosion monitoring of laboratory
specimens and field research are discussed. A new phase of the project involving the use of
corrosion inhibitors prior to wrapping with FRP composites is also described.

xi

CHAPTER 1: INTRODUCTION
1.1

INTRODUCTION TO CORROSION IN REINFORCED CONCRETE BRIDGES

In 1998, the Federal Highway Administration (FHWA) rated 29.6% of the nations bridges as structurally
deficient or functionally obsolete. While this number of problem structures is unacceptable, it has been
declining over the last decade from a high of 34.6% in 1992. The FHWA strategic plan calls for less than
25% of the nations bridges to be classified as deficient by 2008 (ASCE 2001).
With the increased demand on our nations bridges and limited funding towards infrastructure, replacing
deficient structures is not always a viable option. In the 21st century, engineers will increasingly need to
design retrofits in order to preserve the infrastructure already in place.
The economic loss due to problems associated with corrosion of the steel reinforcing in reinforced
concrete is estimated to be $50 billion per year in the United States. It is the single most expensive
corrosion problem in the nation. It affects the integrity of thousands of bridges, roadbeds and overpasses
(Jones 1996).
Concrete is typically a strong, durable, and long lasting building material. However, in an aggressive
environment, a concrete structure may experience premature deterioration due to exposure to salts. While
it may be expected that structures in a marine environment will be exposed to salts, structures located
above the freezing line are also exposed. This is due to the policy that started in the 1960s of applying
deicing salts to roadways and bridge decks in order to keep them clear of ice. As a result, many bridges
across the country have been experiencing problems associated with corrosion. The exposure to salt and
water results in damage to the structure in the form of cracking, delamination, and spalling. This damage
may cause structural problems and requires expensive rehabilitation. Overall, developing innovative ways
to prevent corrosion from taking place and implementing long-term solutions to repair chloride
contaminated concrete are necessary endeavors.
A recent solution for repairing damage due to corrosion in reinforced concrete is to use fiber reinforced
plastic (FRP) composite wrap. In addition to strengthening a concrete member, the FRP wrap is not
subject to corrosion and may provide a barrier to protect the concrete from an aggressive environment. A
potential concern is that by encapsulating concrete with FRP composites, moisture, chlorides, and oxygen
may be trapped and cause the corrosion process to continue undetected. Figures 1.1 and 1.2 show a bridge
bent before and after repair.

Figure 1.1 Corrosion damage to a bridge bent

Figure 1.2

1.2

Bridge bent and column wrapped with FRP after

repair of corrosion damage

PREVIOUS RESEARCH

TxDOT Project 1774, The Effect of Wrapping Chloride Contaminated Structural Concrete With Layers
of Glass Fiber/Composites and Resin, was initiated in 1997 in order to address questions regarding the
effectiveness of FRP composites for corrosion related repairs.
There are several ongoing field studies by other organizations that will also provide researchers
information on the performance of FRP composite wrap in the repair and prevention of corrosion in
2

reinforced concrete. There are not any significant findings to date from these field studies. Findings from
previous laboratory studies will be discussed in Chapter 3. However, these studies rely on extreme
exposures that are necessary for accelerated laboratory simulation of corrosion. While they may be used
to predict behaviors that might happen in the field, they do not accurately reflect conditions that are found
in the field.
In a study at the University of Sherbrooke, twelve circular columns were rehabilitated in 1996 after
experiencing significant damage from corrosion. The columns were located on a bridge on Highway 10,
near Sherbrooke, Quebec, Canada. Five of the columns were wrapped with glass FRP, four were wrapped
with carbon FRP, and three were repaired with conventional repair materials. Existing chloride levels in
the structures were not addressed other than to remove and repair the delaminated and spalled concrete
prior to wrapping. Fiber optic sensors were installed to monitor axial deformations and circumferential
expansion in each of the columns. Many of the sensors were damaged during the wrapping process.
Results from this study are still pending (Labossiere, Neale, and Martel 1997).
In a study at Purdue University, sponsored by the Indiana Department of Transportation, a field
evaluation was performed in addition to a laboratory study on monitoring long-term performance of
highway bridge columns retrofitted with advanced composite jackets. The concrete columns of a bridge at
the junction of Interstate 69 and U.S. Highway 14 in Fort Wayne, Indiana were repaired with glass FRP in
1997 after experiencing deterioration due to corrosion and extreme temperatures. There was no mention
in the report as to what other repairs were made to the columns, or if chloride contamination of the
concrete was addressed. Thermocouples and strain gages were installed on three of the columns. Visual
inspection was performed periodically over two years. Damage to the FRP wrap was observed due to auto
collisions, and the damaged areas had increased during subsequent inspections. The increase in damage to
the wrap was speculated to be caused by the glass fibers sensitivity to moisture. No other findings from
the field evaluation were reported (Teng, Sotelino, and Chen 2000).

1.3

PROJECT OBJECTIVES

The investigation for TxDOT Project 1774 consists of both a laboratory research program and a field
research program. The objective is to determine the long-term effectiveness of repairs with FRP
composite wrap. The laboratory study consists of reinforced concrete specimens that are subjected to an
accelerated corrosive environment. Both chloride contaminated concrete and non-chloride contaminated
concrete are being evaluated. The field study consists of long-term monitoring of highway bridges that
were wrapped with FRP. Corrosion rate probes were embedded into some of the structures prior to being
wrapped.
This report will discuss the mechanisms of corrosion of steel embedded in concrete, the use of FRP
composites, and corrosion monitoring techniques. The short-term findings from monitoring of both
laboratory and field studies will be discussed and compared. Also, the initiation of an additional
laboratory study on the use of corrosion inhibitors in conjunction with FRP composite wrap will be
discussed.

CHAPTER 2: CORROSION OF STEEL IN CONCRETE

2.1

MECHANISMS OF CORROSION

Corrosion is defined as the destructive result of chemical reactions between a metal or metal alloy and its
environment. The process involves a transfer of an electronic charge in an aqueous solution (Jones 1996).
In steel, corrosion results in the metallic iron being converted to the voluminous corrosion product ferric
oxide (Mailvaganam 1992). Reinforcing steel is susceptible to uniform corrosion and pitting that may be
caused by several different mechanisms as will be discussed later.
New concrete typically has a pH of 12.5 or higher. The high alkalinity passivates the steel by forming a
protective film of ferric oxides. This film prevents the reinforcing steel from corroding. However, certain
factors can increase the likelihood of the passive film being broken down. These factors include
development of cracks in the concrete that extend to the steel, high permeability and/or high porosity of
the concrete, inadequate concrete cover over the steel, or high levels of chlorides (Perkins 1997).
Once the passive film breaks down, an electrochemical reaction takes place. This reaction may be
separated into two partial or half-cell electrochemical reactions. The first reaction is the oxidation reaction
or anodic reaction [2.1]. The second reaction is the reduction reaction or cathodic reaction [2.2]. The two
half-cell reactions are expressed below:
The anodic reaction: Fe Fe2+ + 2eThe cathodic reaction: 2e- + H2O + O2 2OH-

[2.1]
[2.2]

This leads to the development of ferric oxide or rust. The formation of ferric oxide is expressed in the
following equations (Broomfield 1997):
Ferrous hydroxide: Fe2+ + 2OH- Fe(OH)2
Ferric hydroxide: 4Fe(OH)2 + O2 + H2O 4Fe(OH)3
Hydrated ferric oxide: 2Fe(OH)3 Fe2O3H2O

[2.3]
[2.4]
[2.5]

The passivated area becomes the anode, and the remaining steel becomes the cathode. The moist concrete
acts as an electrolyte, allowing the electrons to move from the anode to the cathode (Allen, Edwards, and
Shaw 1993). The corrosion process is illustrated in Figure 2.1.

Figure 2.1

The corrosion process (Broomfield 1997)

5

When iron oxides are formed, they have a volume that is between two to ten times greater than the
volume of the steel that it replaced. This volume increase places considerable tensile stresses on the
concrete, which leads to cracking of the concrete (Broomfield 1997). After cracks have formed, the
concrete may begin to separate at the level of the reinforcing along a plane parallel to the concrete
surface. This separation is referred to as delamination. As the corrosion process continues, pieces of
concrete may eventually break away or spall off. This process is illustrated in Figure 2.2: (a) is the
exposure to corrosive environments, (b) is the formation of cracks, and (c) is the formation of
delamination and spalling (Scannell, Sohanghpurwala, and Islam).

Figure 2.2

Consequences of corrosion of steel in concrete.

There are two main causes of corrosion of steel in reinforced concrete: carbonation of the concrete and
chloride attack. Carbonation results when carbon dioxide in the air reacts with calcium hydroxide in the
concrete, which causes a significant reduction of the pH. The carbon dioxide gas dissolves in the pore
water to form an acid. The acid is then neutralized by the calcium hydroxide, which forms calcium
carbonate. The pH level in the concrete drops as the calcium hydroxide reacts with the carbon dioxide
(Kay 1992).

Carbonation is most common in concrete with a high water-cement ratio. It will most likely be found near
the concrete surface and adjacent to cracks. The rate of penetration is proportional to the square root of
time; therefore it slows down with increased depth (Allen, Edwards, and Shaw 1993). Carbonation does
not significantly affect the strength of the concrete, but it will affect the steel if allowed to penetrate that
far. The pH level of carbonated concrete may drop below 8; hence adjacent steel cannot maintain its
passive film and will proceed to corrode (Broomfield 1997).
Chloride attack, which results in pitting, takes place when the concrete is exposed to high levels of
chloride ions. Concrete may be exposed to chlorides through a number of different sources. The most
common are exposure to deicing salts or seawater. Some other sources of chlorides include the use of
chloride containing additives in the concrete mix or the use of chloride contaminated water or aggregates
(Hansson and Sorensen 1990). The chloride ions cause very localized breakdowns in the passive film of
the steel. Hence a small anode is formed and the rest of the passive film serves as the cathode. Due to the
large cathode-to-anode ratio, the rate of attack is accelerated and a pit is formed (Mailvaganam 1992).
Chloride thresholds are approximate due to variations in pH level of the concrete, amount of chlorides
that are bound chemically or physically, and the amount of moisture found in the concrete. The most
commonly quoted threshold is about 1 lb of chloride per cubic yard of concrete, which is approximately
0.03% chlorides by weight of concrete and 0.22% chlorides by weight of cement, depending on the
concrete mix. In laboratory tests (Broomfield 1997), the chloride threshold was found to be 0.4% chloride
by weight of cement for cast-in-chlorides and 0.2% for diffused-in-chlorides. Table 2.1 lists the American
Concrete Institutes recommended limits for chloride ion content in new structures (Gaynor 1987).

Table 2.1

Recommended limits for water-soluble chloride ion

content in concrete, from ACI 201.2R-77
Maximum water-soluble chloride ion
content, percent by weight of cement

Category of concrete service

Prestressed concrete

0.06

Conventionally reinforced concrete in a

moist environment and exposed to chloride

0.10

Conventionally reinforced concrete in a

moist environment but not exposed to
chloride

0.15

Above ground construction where concrete

will stay dry

No limit for corrosion

Cracking greatly increases the corrosion process. Carbonation is typically found along cracked surfaces,
and penetrates into the concrete adjacent to the crack. Steel in carbonated concrete is more susceptible to
chlorides since it is no longer passivated. Cracks that extend to the steel reinforcing provide a direct path
for chlorides, moisture, and oxygen which creates an environment ideal for corrosion to take place
(Mailvaganam 1992).

2.2

PREVENTION OF CORROSION

In order to prevent corrosion, care should be taken in the mixing and pouring of concrete. The concrete
mix should have a low water-cement ratio to reduce permeability. The aggregate should be well graded in
order to ensure a good adherence between the cement and aggregate. The concrete should be well

consolidated when it is cast in order to prevent voids from forming. Care should also be taken to keep a
minimum of 2 in. cover between the surface of the concrete and the reinforcing steel (Jones 1996).
In addition to quality workmanship, engineers have developed other solutions to reduce the possibility of
corrosion. One solution is the use of corrosion inhibitors. Other solutions include modifying the
reinforcing by using epoxy or zinc (galvanized) coatings. Another possibility is to use FRP bars instead of
steel.
Corrosion inhibitors, such as calcium nitrite based additives, may be added to the concrete mix or applied
to the concrete surface. They create a thin chemical layer at the steel surface, which inhibits corrosion.
There is concern as to how well corrosion inhibitors migrate to the steel when applied to the surface, and
whether they are able to protect against localized pitting (Broomfield 1997). The use of corrosion
inhibitors will be discussed in greater detail in Chapter 8.
Another method of preventing corrosion of the steel is to limit the ability of the reinforcement to corrode
by using protective coatings or FRP instead of steel. Such reinforcement costs more than traditional black
steel. The characteristics and behavior of such reinforcement continue to be investigated by researchers.
The use of fusion bonded epoxy coated reinforcing for corrosion protection has been around since the
1970s. It is the most popular of the corrosion protection strategies, and is used in forty-one states in the
United States (Babaei and Hawkins 1991) Epoxy coating improves the performance of reinforcement in a
corrosive environment provided that coating damage is minimized. Defects in the coating that may result
from improper fabrication or handling practices can cause a significant decrease in corrosion protection.
Epoxy coated reinforcement has a lower ultimate bond strength than black steel that must be taken into
account in the design.
Galvanized steel reinforcement has been in use since the 1930s. The zinc coating can remain passive in a
lower pH level and can withstand higher chloride concentrations than black steel. Although the ultimate
bond strength is often found to be nominally higher than black steel, the passivation of the zinc generates
hydrogen, which may reduce the bond strength (Yeomans 1991). Galvanized steel reinforcement is not
commonly used in the United States. The Federal Highway Administration (FHWA) recommends that
fusion bonded epoxy coated reinforcing be used instead of galvanized steel because galvanized steel is
more susceptible to chloride attack. However, prevention of coating damage in galvanized steel is not as
crucial due to the sacrificial nature of the zinc coating (Broomfield 1997).
FRP reinforcement has been in use since the 1980s. Over a dozen prestressed concrete pedestrian and
roadway bridges have been built, primarily in Japan and Germany. These bridges use FRP cables instead
of steel prestressing tendons. Unlike steel, FRP reinforcement does not corrode. Since it is a new material,
the long-term strength, durability, and fatigue limits have not been resolved. Also, special attention must
be paid to anchorage details, since FRP behaves differently than steel (Erki and Rizkalla 1993). The use
of FRP reinforcing bars instead of steel reinforcing bars is an even newer development. The first roadway
concrete bridge deck to have FRP reinforcing bars was built in West Virginia in 1996. FRP reinforcing
bars have additional design issues because the physical properties of the reinforcement vary from
manufacturer to manufacturer. Therefore design calculations must be made for the specific reinforcement
used (Bassett 1998).

2.3

REPAIR METHODS

Common repair techniques for corrosion damage involve removing the cracked and spalled concrete over
and around the reinforcing bars, cleaning or replacing badly corroded reinforcement, and replacing the
concrete removed with a patch. This method is tedious and often the repairs need to be made again in less
than two years (Watson 2000).
When patching concrete, macrocell corrosion may form due to physical, chemical, and electrochemical
dissimilarities between the existing concrete and the patch material. This is due to localized corrosion
cells between the reinforcing in the existing chloride contaminated concrete and that in the new, patched

area. The localized corrosion, referred to as a macrocell, accelerates the corrosion process in the repair
area. This often leads to spalling of the patch (Whitmore, Abbott, and Velivasakis 1999).
To minimize the possibility of macrocell corrosion, it is necessary to remove all of the carbonated,
chloride contaminated, and other unsound concrete. Then the steel reinforcing must be thoroughly
cleaned or replaced to remove all corrosion products. Finally, the repair material should be similar to the
existing sound concrete, to reduce electrochemical differences due to different oxygen permeabilities of
the two materials (Gu, et al. 1997).
Other techniques used for repair and prevention of corrosion involve use of latex modified mortars in
repair patches, injection of cracks with polymer resins, cathodic protection of reinforcement, chloride
extraction, and more recently the use of fiber reinforced plastics (Perkins 1997).
Polymers have been used in cementitious mortars since the 1950s. Admixtures containing latex polymers
may replace a portion of the mixing water in the repair mortar. It functions as a water reducing plasticizer,
which lowers the water-cement ratio and reduces the permeability of the mortar. It also improves the bond
between the patch material and the existing concrete (Allen, Edwards, and Shaw 1993).
Crack injection is primarily used to repair deep cracks and to improve the structural strength of a concrete
member that is compromised due to a crack. In crack injection, epoxy, polyurethane, or polyester resin is
injected into a crack after any corrosion product, dirt, and/or grit is removed. The polymer resin bonds
the concrete across the crack and seals the crack against the ingress of water and reduces the ingress of
carbon dioxide (Perkins 1997).
Cathodic protection reduces the corrosion of a metal exposed to an aqueous electrolyte. An anode, either
impressed current or sacrificial, is connected electrically to the steel reinforcing. This causes a cathodic
polarization of the steel reinforcing, which may reduce or prevent corrosion (Jones 1996).
Chloride extraction is an electrochemical process to move chloride ions away from the steel
reinforcement in chloride contaminated concrete. A voltage is applied to an external mesh that is installed
on the surface of the concrete. The mesh becomes a positively charged anode and attracts the negatively
charged chloride ions. Hydroxyl ions are produced at the negatively charged steel reinforcement, which
causes the pH to increase and the concrete to become more alkaline (Kay 1992).
The use of fiber reinforced plastics for repair of concrete and prevention of corrosion will be discussed in
the following chapters.

10

CHAPTER 3: FIBER REINFORCED PLASTIC COMPOSITES

3.1

INTRODUCTION TO FRP COMPOSITES

Fiber reinforced plastic (FRP) composites consist of fiber reinforcement that is suspended in a polymer
resin matrix. The fibers provide increased stiffness and tensile capacity, and the resin offers high
compressive strength along with binding the fibers together. Three different types of fibers: glass, carbon,
and aramid may be used depending on application and budget. Glass is the least expensive of the fibers.
Carbon is considerably more expensive, and aramid is the most expensive. Glass, carbon, and aramid FRP
composites are abbreviated GFRP, CFRP, and AFRP respectively. Table 3.1 includes properties of the
different types of FRP as well as steel.

Table 3.1 Basic Properties of Glass, Carbon, and Aramid FRP

Compared to Steel (Bassett 1998)
Property
Tensile strength (ksi)
Modulus of elasticity (Gsi)

Glass
FRP

Carbon
FRP

Aramid
FRP

Steel

200-250

240-350

170-300

200-270

7-9

22-24

7-11

27-29

0.03-0.45 0.01-0.015 0.02-0.26

Elongation
-6

0.04

Coefficient of thermal expansion (10 /F)

5.5

0.0

-0.5

6.5

Specific gravity

2.4

1.6

1.25

7.9

Glass fibers are divided into three classes: electrical, structural, and corrosion resistance. Electrical,
E-glass, is the most commonly used for civil engineering applications. It is the most economical, and also
is an excellent electrical insulator. Structural, S-glass, has a higher strength and corrosion resistance than
E-glass. Corrosion resistant glass, C-glass, has better resistance to acids and bases than S-glass, but is not
commonly used in civil engineering. Glass fibers creep under sustained loads, and degrade under
increased temperature. These factors have to be taken into consideration in the design (Tang 1997).
Carbon fibers are made from polyacrylonitrile (PAN), pitch, or rayon polymer precursors. They are
available in bundles of untwisted carbon filaments, referred to as tow. They are available in several
different modulus categories ranging from standard (33-35 Msi) to ultrahigh (50-70 Msi). Carbon fibers
have a very high fatigue and creep resistance and are more brittle than glass or aramid fibers. Carbon
fibers are also susceptible to galvanic corrosion when placed next to a metal (Bassett 1998).
Aramid fibers are synthetic organic fibers consisting of aromatic polyamides. The commercial grade of
aramid fibers is known as Kevlar. They have excellent fatigue and creep resistance. When used in a
composite, the fibers can have difficulty achieving a chemical or mechanical bond with the resin. Aramid
fibers also have a low compressive strength, so they are not well suited for applications involving high
compressive loads (Mallick 1997).
Structural applications of composites generally use long, continuous fibers, as opposed to short,
discontinuous fibers that are for applications such as automobile bodies. The most common forms of
fibers are rovings, tows, and fabrics. Rovings are similar to tows, except they are bundles of glass or
aramid fibers, rather than carbon fibers. Fabrics consist of small bundles of fibers that are woven together
to provide strength in both the 0 (warp) direction and the 90 (weft) direction (Bassett 1998).

11

There are three different methods for manufacturing commercial composite products: pultrusion, filament
winding, and hand layup. Pultrusion is the process of pulling fibers through a resin bath and then through
a heated die to produce a structural shape with a constant cross section. Pultrusion is used to manufacture
rods, beams, and channels. Filament winding is the automated process of winding resin saturated fibers
around a mandrel to produce a circular shape. This method is primarily used to manufacture tanks, pipes,
and poles, but has also been used as an automated method to encapsulate bridge columns. Hand layup
involves saturating sheets of fabric with resin by hand or machine, and then applying them by hand to an
existing structure (Ballinger 1991). This method is primarily used to produce laminates and wraps that are
used in retrofits. Figure 3.1 is an example of the hand layup method being used in the field.

Figure 3.1 FRP composite wrap being applied to the bent of a bridge
Fibers typically occupy 30 to 70% of the matrix volume in a composite. The rest is occupied by the resin
system. Resin systems consist of two classes, thermoplastics and thermosets. Thermoplastics are not used
in structural applications because they do not cure permanently. The polymers in a thermoset resin
irreversibly cross link, which causes the resin to remain solid at elevated temperatures after it has cured.
The most common resin systems used are unsaturated polyesters, epoxies, and vinyl esters (Tang 1997).
Unsaturated polyesters account for the majority of the resins used. Their quick curing rate and affordable
cost makes them ideal for pultrusion. Epoxies are typically the most expensive resins. They are generally
used for applications that require high performance. The high viscosity of epoxy resins limits their use to
filament winding and hand layup manufacturing. Vinyl esters have the workability of epoxies and the fast
curing rates of polyesters. They have higher physical properties than polyesters, and cost less than epoxies
(Mallick 1997).
The research study discussed herein involves fiber composite wraps. There are many advantages of using
FRP composites over steel. The strength-to-mass density ratio is 10 to 15 times higher than steel. Overall,
FRP materials have very high strength-to-weight and stiffness-to-weight ratios. These properties allow
FRP composites to be used to strengthen a structure without adding any significant additional weight to
the structure. They are also corrosion resistant and have a low axial coefficient of thermal expansion.
These characteristics offer advantages over steel in extreme environments (Erki and Rizkalla1993).
Prior to the development of externally applied FRPs, steel jackets were often used. Steel jackets increase
the stiffness, strength, and energy absorption of the concrete, along with enhancing ductility and
providing additional shear strength (Mirmiran and Shahawy 1997). A major drawback of steel jackets is
12

the installation. It is very difficult to fit steel jackets and requires the use of heavy clamping tools and
welding, resulting in high installation costs. Also steel jackets are not practical in an aggressive
environment due to their susceptibility to corrosion.
Unlike steel jackets, FRP materials are easy to handle, can conform to the shape of existing elements, and
can be applied quickly. In the case of retrofitting, this means that the repair time will be shorter than other
methods of repair. Shorter repair times result in savings in labor and construction related costs and
benefits the public by reducing the time a structure must be closed. When calculating the overall cost
effectiveness and lifetime expectancy for the repair, the advantages of FRP composites may offset the
high initial cost (Bassett 1998).
There are several disadvantages that have limited the use of FRP composites in the field of civil
engineering. From the point of view of material properties, high cost, low modulus of elasticity, low
failure strain, and little information about long-term durability are prime disadvantages. Another
disadvantage is that FRP composites are new materials, and engineers are unfamiliar with their properties
or applications. For implementation in infrastructure applications, a drawback is the lack of predictive
models and design guidelines (Karbhari, Seible, and Hegemier 1996).
In a study performed at the University of Central Florida, the behavior of FRP encapsulated columns was
compared to different confinement models. The results were that the use of FRP significantly increased
the strength and ductility of the concrete. The researchers also found that the characteristics of
confinement using FRP are different than the characteristics of confinement using steel. Steel is an
elastoplastic material and FRP materials are linearly elastic, but most models for confinement do not
account for this. The study showed that many confinement models overestimated the strength of FRP
encapsulated concrete. Until models based on FRP are developed, the factor of safety for design of
confinement in FRP encapsulated concrete should be increased. A higher factor of safety will decrease the
cost effectiveness of FRP materials (Mirmiran and Shahawy1997).

3.2

SEISMIC APPLICATIONS

FRP materials have already been established as a method for strengthening concrete and preventing
seismic damage. Studies in the last decade have shown that wrapping FRP fabric around the perimeter of
concrete columns improves ductility, strength, and confinement of the column (Toutanji and Balaguru
1998).
Concrete structures built prior to the 1970s are expected to have the worst response in the event of an
earthquake. This is due to the poor detailing resulting in a lack of ductility in reinforced concrete
members. Concrete with poorly detailed reinforcement has little capacity to absorb the energy imposed on
the structure during seismic loading. Changes to the 1976 Uniform Building Code increased the ductility
requirements and structures designed using recent codes have performed much better in recent
earthquakes (Feld and Carper 1997). Figure 3.2 is an example of a nonductile failure.
By wrapping columns with FRP, the high tensile strength of the composite increases the confinement of
concrete in the column. The confinement provided by the FRP composite causes the concrete to fail at a
larger strain thereby increasing the ductility of the column. The lateral restraint provided by the FRP
increases the compressive strength of the concrete confined within the wrapping and results in higher
load-carrying capacity of the column. The lateral confinement also provides additional restraint for
buckling of the longitudinal reinforcing. FRP composites have the flexibility to be wrapped around either
circular or rectangular columns, although sharp corners must be avoided. The low weight and negligible
increase in stiffness of FRP composites offers advantages over steel jacketing of columns (Saadatmanesh,
Ehsani, and Li 1994).

13

Figure 3.2 Highway bridge column after the 1994 Northridge

earthquake (Courtesy of EERC, University of
California, Berkeley).
Wrapping technology received considerable attention in Japan. Columns were strengthened by externally
bonding sheets of carbon fibers to enhance resistance to seismic loadings. After the superior performance
of carbon FRP encapsulated columns in the 1995 Kobe earthquake in Japan, use of the technology
increased further. As of 1996, over 1000 column strengthening projects in Japan have used FRP
composite wrap (Thomas, et al 1996).
In the United States, 15 columns on the bridge ramps of Interstate 5 and Highway 2 in Los Angeles,
California were wrapped with glass FRP composites in 1991 as part of a seismic retrofit program initiated
by the California Department of Transportation. The columns were 6 feet in diameter and ranged from 18
to 55 feet in height. They were located 15 miles southwest of Northridge. The columns showed no
damage following the 1994 Northridge earthquake. After the Northridge earthquake, the use of FRP
composite wrap has steadily increased in California (Fyfe 1995).

3.3

APPLICATIONS IN CORROSION REPAIR AND PREVENTION

Since FRP composite materials are not susceptible to corrosion, engineers have been exploring the
possibilities of encapsulating concrete with FRP instead of steel jackets. Currently FRP composite wraps
are being investigated for strengthening of bridges that are deteriorating due to corrosion and also to
prevent future corrosion activity.
Watson recommends the use of FRP composite wrap for non-seismic repairs and upgrades of highway
bridges. One of the advantages he cites in using FRP composite wrap is that it Protects the concrete with
an impervious barrier that will prevent de-icing chlorides from attacking the steel rebar, (Watson 2000).
In the second edition of Construction Failures, the authors Feld and Carper state that applying an
impervious coating over contaminated concrete is not a valid solution to repairing corrosion damage due
to chlorides (Feld and Carper 1997). Leeming and Peshkam, in their analysis of the advantages of FRP
plates, state that one of the advantages of FRP composites over steel plates is that it is not necessary to
remove all chloride contaminated concrete because the composite material would stifle any further
corrosion (Leeming and Peshkam 1995). In another analysis of the advantages of FRP over steel plates by
Emmons, Vaysburd, and Thomas, the authors disagree with Leemings and Peshkams conclusion that
FRP encapsulation will stop the corrosion process. They state that unless the corrosion problem has been
properly determined and addressed, encapsulation will most likely accelerate the corrosion process in the
existing structure. (Emmons, Vaysburd, and Thomas 1998)
14

Overall, there has been very little research in the long-term effect of FRP composite wrap on the
prevention of corrosion of steel reinforcing in concrete. In an investigation by Ohta, et al., they evaluated
the rehabilitation of deteriorated prestressed concrete beams. The bridge was a three-span post-tensioned
prestressed concrete bridge with simple tee-beams, built in 1957, and located near the Sea of Japan. In
1977, the bridge was rehabilitated after showing signs of severe cracking. The cracks were injected with
an epoxy grout. The beams experiencing cracks were strengthened with a 4.5-mm steel plate bonded to
the lower flange and two layers of GFRP composite wrap were applied to the web. The rest of the beams
had one layer of GFRP applied to the entire surface. The rehabilitation was evaluated in 1990 when the
bridge was scheduled to be demolished. The researchers found that the FRP prevented further chlorides
from penetrating the concrete. They also found that the deterioration of the bridge due to corrosion of the
steel reinforcing bars and sheaths had continued due to insufficient measures that were taken to address
the chloride contamination of the concrete at the time of rehabilitation (Ohta, et al. 1992).
On the Bryant Patton Bridges in Florida, it was found that fiberglass jackets applied to bridge piles in a
marine environment had accelerated the corrosion process. Capillary action allowed moisture to rise from
the submerged portion of the pile, but the fiberglass jacket prevented the concrete from drying out. The
combination of moisture and the high levels of chlorides in the unrepaired portions of the pile resulted in
severe damage due to corrosion, as seen in Figure 3.3 (Sohanghpurwala and Scannell 1994).

Figure 3.3 Concrete pile that had previously been repaired with a fiberglass jacket.
In a laboratory study at Purdue University previously mentioned in Chapter 1, researchers concluded that
FRP composite materials provide excellent protection against aggressive environmental conditions. In the
study, five 20 in. x 10 in. reinforced concrete beams were exposed to 56 weeks of wet/dry cycles of 5%
saline water. The specimens consisted of new concrete; there was no attempt to replicate conditions that
might be found in damaged concrete. One beam was unwrapped, one beam was coated with epoxy, one

15

was wrapped with one layer of GFRP, and two beams were wrapped with two layers of GFRP. At the end
of the study, the unwrapped specimen was the only one that showed signs of corrosion activity (Teng,
Sotelino, and Chen 2000).

3.4

LONG-TERM BEHAVIOR

The long-term behavior of FRP composite materials used in infrastructure is not fully understood. One
concern is whether the long-term durability of the material will remain the same, or if the effects of
moisture, freezing environment, and/or ultraviolet light will cause the material to degrade and
compromise the expected performance of the material. Most composite materials have not been in service
long enough to accurately evaluate their durability.
Brandt Goldsworthy reports that composites will have a service life considerably longer than traditional
materials. He cites that the strength of composite materials used in submarines and electric utility poles
was 5 to 7% higher after 30 years of service. The increase in strength is due to the continuous curing of
the resin systems. Goldsworthy also states that the UV degradation of composites is self-limiting. After
the surface oxidizes, it protects the rest of the material from further degradation (Bassett 1998).
In the aforementioned investigation by Ohta, et al., they found the tensile strength of the FRP was reduced
by 60% after 13 years of exposure. In the load tests on the beams, the overall contribution of the
rehabilitation to the load carrying capacity was significantly smaller than the degree expected.
A study performed jointly at the University of Alabama and Rutgers University, Toutanji and Balaguru,
examined the durability of carbon and glass FRP materials externally applied to concrete and subjected to
different environments. The environments used in the study were room temperature, wet/dry cycles using
salt water, and freeze/thaw cycles.
Twenty-four concrete cylinders that were 305 mm long and 76 mm in diameter were wrapped with two
layers of FRP material. The different parameters were unwrapped cylinders, cylinders wrapped with
carbon FRP (CFRP), and cylinders wrapped with glass FRP (GFRP). Each parameter was subjected to
different environments. One set was kept at room temperature for 75 days. Another set was exposed to
3000 wet/dry cycles using 3.5% sodium chloride solution. The cycle consisted of 4 hours wet followed by
2 hours dry. The last set was exposed to 300 freeze/thaw cycles. The cycles ranged in temperature from
4.4C to 17.8C within 4 hours. After the exposure cycles were completed, the specimens were loaded
axially until failure.
The results indicated that for the CFRP wrapped cylinders in a room temperature environment, the
compressive strength increased by 200%. GFRP wrapped cylinders in this same environment showed a
100% increase in compressive strength compared to the unwrapped specimens. In the wet/dry
environment, the wrapped cylinders experienced less than 10% reduction in strength. In the freeze/thaw
environment, the wrapped specimens experienced a 20 to 28% reduction in strength while the unwrapped
specimen disintegrated.
Overall, the FRP encapsulated concrete showed an increase in strength over the unwrapped concrete. The
ductility of the GFRP cylinders decreased considerably after exposure to harsh environments. The
wrapped cylinders exhibited more catastrophic failures after exposure to the freeze/thaw cycles than the
other environments (Toutanji and Balaguru 1998).
In a similar study by Soudki and Green, the behavior of columns wrapped with CFRP was investigated.
The column specimens were 150 mm by 300 mm cylinders, half of which were reinforced with steel, and
the other half plain concrete. The parameters were unwrapped, or wrapped with one or two layers of
CFRP. The cylinders were exposed to 200 freeze/thaw cycles, and then subjected to axial compression
tests. The cycles ranged from 18C to 18C within 24 hours. They found that one layer of CFRP
increased the strength up to 57%, compared to unwrapped specimens exposed to the same environment.
The second layer of CFRP increased the strength by an additional 30%. They also found that the

16

specimens exposed to freeze/thaw cycles failed in a more catastrophic manner than the specimens that
were not exposed (Soudki and Green 1997).
The Ministry of Transportation in Ontario is currently sponsoring an ongoing research project by The
University of Toronto on using FRP materials to repair damaged chloride contaminated concrete columns.
The purpose of the study is to test the effectiveness of repairing damaged concrete with different types of
grout and then wrapping with GFRP composites (Sheikh, et al. 1999).
The study involved subjecting reinforced concrete columns, 1524 mm long and 406 mm in diameter, to
accelerated corrosion. A steel tank was installed around each column and then filled with a 2% sodium
chloride solution. The steel reinforcing was connected to a voltage source in order to form a corrosion
cell. The columns were subjected to accelerated corrosion for seven months, resulting in light to moderate
corrosion of the longitudinal reinforcement and more severe corrosion of the spiral reinforcement. The
damaged concrete was removed and then the columns were repaired using different mortars and polymer
coatings. After the repairs were made, the columns were wrapped with two layers of GFRP. The repaired
columns, along with an undamaged column and an unrepaired column, were each subjected to an axial
load until failure. The results of the tests found that the maximum load carried by the columns repaired
with FRP composite was similar in value to the maximum load carried by the column that had not
experienced damage due to corrosion. The FRP encapsulated columns were more ductile than the
undamaged column. The unrepaired column failed in an explosive manner rather than a ductile failure
(Sheikh, et al. 1999). In this study, the FRP encapsulation compensated for the loss of steel resulting from
corrosion.

17

18

CHAPTER 4: CORROSION MONITORING METHODS

4.1

INTRODUCTION TO CORROSION MONITORING

Monitoring of corrosion activity is essential for establishing preventive maintenance and repair
procedures. Monitoring may range from visual observation to taking measurements to assess presence or
rate of corrosion. Visual observation is often used to assess whether corrosion is the cause of damage.
More sophisticated methods may be used to determine whether corrosion is taking place before the
concrete becomes damaged, and to assess the corrosion rate.

4.2

VISUAL INSPECTION

Visual inspection is used to determine the condition of a concrete structure. An inspector can assess to
what extent a structure is damaged, and the cause of the damage. It is the simplest method to determine if
concrete damage is the result of corrosion. A visual inspection is conducted by examining crack patterns,
looking for rust stains and spalling, and checking for delamination of the concrete cover (Scannell,
Sohanghpurwala, and Islam 1996).
Mapping crack patterns and determining crack size is one of the first steps in a visual inspection. Cracks
resulting from the formation of rust are generally parallel to the length of the reinforcing and are a good
indication that corrosion activity is present. Although any rust stains should be noted, they are not as
reliable as an indication of corrosion. Rust stain may be caused by contamination of the aggregates with
iron pyrites (Allen, Edwards, and Shaw 1993). The size and location of concrete spalls should be noted,
since they reflect serious distress.
A hammer survey is the most cost effective method to determine the extent of delamination. It is quicker
and less expensive than methods involving radar, ultrasonic, or infrared thermography, and often more
accurate (Broomfield 1997). However, new, less expensive, more mobile instruments are likely to be
available in the future. A hammer survey is conducted by tapping concrete with a medium weight
hammer and listening for a hollow sound. The hollow sounding areas are then marked as areas of
delamination.
The presence of carbonation or chlorides can be determined by simple field tests. Carbonation is
identified by spraying freshly exposed concrete with a solution of phenolphthalein in diluted ethyl
alcohol. The solution changes from colorless to purple-pink as the pH rises above 10. Carbonated
concrete will remain colorless and uncarbonated concrete will become stained. There are several chemical
tests that can be performed on site to detect chlorides. A sample of powdered concrete is collected and
dissolved in a chemical agent. Special indicator paper is then dipped into the solution to determine the
percentage of chlorides by mass of concrete (Allen, Edwards, and Shaw 1993).
While visual inspection is useful to assess the overall structure, it is difficult to determine the onset of
corrosion activity.

4.3

HALF-CELL POTENTIAL

The use of halfcell potential measurements for corrosion testing was developed in the 1960s and is
covered by ASTM Specification C876. Although it does not determine the rate of corrosion, it does
estimate the likelihood of corrosion taking place. Since the corrosion process causes an electrical potential
to be generated, the half-cell provides a reference against which the corrosion potential can be measured.
The corrosion potential, also known as Ecorr, can be compared to criteria established through laboratory
testing.
The half-cell consists of a reference electrode of a metal contained in an electrolyte solution saturated
with one of its own salts. The reference electrode is in a rigid plastic or glass tube with a porous plug. The

19

most commonly used reference electrodes are copper/copper sulfate (Cu/CuSO4) and saturated calomel
(SCE) (CRC 1991). Since the ASTM Specification C876 is for a Cu/CuSO4 reference electrode, if a SCE
reference electrode is used instead, the potentials should be converted to Cu/CuSO4 equivalent potentials
(ASTM 1989).
The test is conducted by connecting the reference electrode to the negative terminal of a high impedance
voltmeter. A direct electrical connection is made from the steel reinforcing to the positive terminal of the
voltmeter. Reinforcement is usually electrically continuous in a structure so that only one electrical
connection is needed. A wetted sponge at the end of the reference electrode forms a liquid bridge between
the cell and the concrete. A schematic for the half-cell potential test is shown in Figure 4.1.

Voltmeter

Reference
electrode

Connection to
reinforcing
steel

Wetted
sponge

Concrete

Figure 4.1 Copper-Copper Sulfate Half-cell Circuitry (ASTM 1989)

Half-cell potential readings are typically taken in a grid pattern, generally at 4 ft. spacing on a bridge
deck. The reference electrode is placed against the concrete until the reading on the voltmeter stabilizes.
The results then can be plotted in an equipotential contour diagram, similar to the example in Figure 4.2.

20

Figure 4.2 Equipotential contour diagram (ASTM 1989)

The diagram illustrates areas that have a high probability for corrosion to be occurring, and areas that are
likely to be passive. Interpretation of Ecorr readings with respect to a copper/copper sulfate reference
electrode is summarized in Table 4.1.
Table 4.1 ASTM interpretation of half-cell readings
Ecorr (mV)

Corrosion Probability

Corrosion Risk

>0

N/A

Invalid data

0 to -200

10% probability of corrosion

Low risk

-200 to -350

Activity is uncertain

Intermediate risk

< -350

90% probability of corrosion

High risk

< -500

> 90% probability of corrosion

Severe corrosion

There are limitations to the effectiveness of half-cell potential. The method may not be used on coated
reinforcing or concrete with coated surfaces. Another drawback is that the potentials are measured at the
concrete surface and not at the reinforcing. Because of this, the potential measurements are very sensitive
to moisture content, thickness of the concrete cover, resistivity of the concrete, carbonation, and chloride
ingress (Ohtsu, Yamamoto, and Matsuyama 1997).

4.4

LINEAR POLARIZATION

Linear polarization, also known as polarization resistance, is a method well suited for measuring the
corrosion rate of metals embedded in concrete. There are several portable corrosion rate measurement
devices available for nondestructive testing of concrete structures. Four of these devices that operate on

21

linear polarization techniques: the PR-Monitor, the 3LP Device, the Gecor Device, and the NSC Device,
were evaluated by the FHWAs Strategic Highway Research Program.
Linear polarization was derived from experimental data that showed that the slope of the linear curve of
applied cathodic and anodic current vs. potential is inversely proportional to the corrosion rate. Figure 4.3
illustrates how linear polarization is derived graphically (Jones 1996).

Figure 4.3 Hypothetical anodic and cathodic polarization curve

This method was developed as a nondestructive method for corrosion monitoring. Before linear
polarization, most corrosion rate data were collected by measuring the weight loss of metal coupons
exposed to an aggressive environment. Not only is linear polarization considerably faster, it is better
suited for corrosion rate testing outside of the laboratory (Jones 1996).
In concrete, linear polarization is performed by polarizing the reinforcing steel with an electric current,
and then measuring its effect on the half-cell potential. A reference electrode is incorporated with a
variable low voltage DC power supply. The system is connected to a microprocessor that runs the
calculations and stores the data (Broomfield 1997).
Overvoltages, c for the cathodic region and a for the anodic region, are related to the corrosion potential
by equation [4.1]. The extent of the linearity depends on the cathodic, c, and anodic, a, Tafel constants
selected. The relationship of the corrosion rate, icorr, in terms of current density to the cathodic and anodic
applied current densities, ic and ia, the Tafel constants, and the overvoltages is expressed in equation [4.2]
(Jones 1996).
c/a = Ec/a - Ecorr

[4.1]

c/a = c/a log ic/a/icorr

[4.2]

The polarization resistance, Rp, is equal to the ratio of the change in potential over the applied current.
The corrosion rate is equal to the proportionality constant B divided by the polarization resistance [4.3]
22

(Jones 1996). The proportionality constant B is a function of the Tafel constants [4.4], and for concrete is
typically from 26 to 52 mV depending on the passive or active condition of the steel. A corrosion rate in
m/yr is calculated from the polarization resistance in equation [4.5], where A is the surface area of the
steel measured. In many other fields, the electrical resistance of the medium that the metal is in can be
neglected when calculating the polarization resistance. The electrical resistance of concrete is significant
enough that measured Rp will be higher than the actual Rp. Therefore, the solution resistance of the
concrete needs to be taken into consideration when calculating Rp (Broomfield 1997).
icorr = B / RP

[4.3]

B = ac / [2.3 (a + c)]
6

X = (11 x 10 B) / (Rp A)

[4.4]
[4.5]

The major advantage of linear polarization is that it is the only method that determines a corrosion rate. A
corrosion rate gives the engineer an indication of the extent of corrosion taking place. One disadvantage
of linear polarization is that the rate is an instantaneous rate, which reflects the conditions at the time the
test was run. Another disadvantage is that the test is sensitive to factors such as humidity and temperature.
Corrosion rate measurements will increase in warm conditions. Since the resistivity of concrete decreases
when the concrete is wet, corrosion rate measurements increase in wet conditions (Scannell,
Sohanghpurwala, and Islam 1996).
Another disadvantage is that assumptions may have to be made about the area of steel being measured. It
is difficult to calculate the exact area being polarized since the impressed current fans out from the
electrode. A guard ring system confines the area of the impressed current, allowing for a more accurate
estimation of the area of steel being measured. Figures 4.4 and 4.5 show schematics of linear polarization
devices (Broomfield 1997).

Figure 4.4 Schematic of linear polarization device (Broomfield 1997)

23

Figure 4.5 Schematic of linear polarization device with a

sensor controlled guard ring (Broomfield 1997)
The set up for linear polarization is similar to that of half-cell potential in that a connection to the
reinforcing must be made. The reference electrode is placed over a portion of concrete where the size of
reinforcing is known. Corrosion rate measurements generally take longer to run than half-cell potential
measurements. Interpretations of values for icorr are summarized in Tables 4.2 and 4.3 (Broomfield 1997).
Table 4.2 Interpretation of Linear Polarization Results
icorr (A/cm2)

Corrosion Level

< 0.1

Passive condition

0.1 to 0.5

Low to moderate corrosion

0.5 to 1.0

Moderate to high corrosion

> 1.0

High corrosion

Table 4.3 Correlation of corrosion rates to rust growth and section loss
icorr (A/cm2)

Rust growth (m/yr)

Section loss (m/yr)

0.1

1.1

0.5

17.3

5.7

1.0

34

11.5

10

345

115

24

4.5

LONG-TERM MONITORING

For long-term corrosion monitoring, embedded monitoring probes may be desired. Typically they are
used in cathodic protection monitoring systems, but they are ideal for any system where the concrete or
reinforcing is not readily accessible, or a large number of readings are needed over a period of time.
One method of embedded probes is to embed a half-cell reference electrode as close to the reinforcing as
possible. Half-cell potential measurements are then taken from lead wires that extend out of the concrete.
An electrical connection to the steel still needs to be made in order to complete the half-cell circuit. By
embedding the reference electrode, the potential measurements are more accurate since they are taken at
the steel rather than at the concrete surface (Ohtsu, Yamamoto, and Matsuyama 1997).
Another type of monitoring probe allows for linear polarization measurements to be made, so the
corrosion rate can be determined. This type of probe consists of a stainless steel auxiliary electrode, a
reference electrode, and a short length of reinforcement steel. This probe is usually encased in a concrete
prism. An electrical connection is made from the probe to the steel reinforcing in the vicinity of the probe.
By using a probe, it eliminates any assumptions about the area of steel being measured since the length of
reinforcing is already known (John, et al. 1995). These probes are usually not affected by temperature
change, but moisture content and changes in chloride levels in the concrete will affect the corrosion rate
measurements (Perkins 1997).

25

26

CHAPTER 5: PROJECT ORGANIZATION

5.1

TXDOT PROJECT 1774

Project 1774 is a long-term monitoring project to evaluate the effectiveness of FRP composite wrap for
the prevention of corrosion in chloride contaminated concrete. The objective is to compare the findings
from laboratory tests to findings from field installations.

5.2

LABORATORY PHASE

The laboratory phase was designed to replicate different parameters that represent worst-case scenarios
found in the field, and to observe their effect on corrosion. The specimens are exposed to an accelerated
corrosion environment, consisting of alternating wet and dry cycles of 3.5% saline water. The following
parameters were chosen:

Specimen geometry

Chloride content in the concrete mix

Cracking condition

Type of repair material

Application of a corrosion inhibitor

Type of wrapping system

Length of wrapped surface

Condition of the surface (wet or dry) when the wrap was applied

5.2.1 Specimen Geometry

The specimens consisted of two shapes, cylinders and rectangular blocks. The cylinders were modeled to
represent bridge columns, and the rectangular blocks were modeled to represent portions of bridge bents
located at points where bridge deck runoff exposes the bent to water containing deicing salts. The
cylinders were 36 in. long and 10 in. in diameter. The rectangular blocks were 36 in. long with a 10 in. by
10 in. cross section. The reinforcement consisted of four #6 grade 60 bars. The transverse reinforcing for
the cylinders consisted of nine -in. plain steel wire circular hoops spaced at 4 in. The rectangular blocks
have -in. plain steel wire that forms three U-shaped stirrups spaced at 10 in. The steel cage was tied
together with metal ties in order to assure electrical continuity. The reinforcement extended 3 in. past the
concrete to allow easy access for making electrical connections to the steel. The specimens had 1 in. of
concrete cover that was maintained by using plastic chairs. Figures 5.1 to 5.4 illustrate the specimen
geometry.

27

3
1 cover

#6 rebar
36

10
10
Cross section

Profile

Figure 5.1 Geometry of cylinder specimens

Figure 5.2 Typical cylinder specimen

28

1 cover

#6 rebar

36

10

10

10
Profile

Cross section

Figure 5.3 Geometry of rectangular block specimens

Figure 5.4 Typical rectangular block specimen

5.2.2 Concrete
A low quality, highly permeable mix was used for the specimens. In order to insure permeability, the
selected water-cement ratio was 0.7. An air entrainment additive was also added to the mix. In order to
simulate chloride contaminated concrete, half of the specimens had cast-in-chlorides added to the mix.
Food grade salt was added to produce a 3.5% saline solution. The concrete mix mixture properties are
shown in Table 5.1.
29

Table 5.1 Concrete mix design

Item

Amount (lbs./yd3)

Cement

393

Coarse Ag. (3/4-in. crushed limestone)

1970

Fine Ag. (Colorado River sand)

1620

Water

275

Total

4258

The percentage of chlorides found in sodium chloride, NaCl, is calculated in the following equations.
(Gaynor 1987).
Molecular weight:

NaCl = 23 + 35.5 = 58.5

Cl as a percentage of molecular weight: 35.5/58.5 = 60.7%

[5.1]
[5.2]

For this mix design, approximately 9.6 lbs. of salt per cubic yd. of concrete were added to make the
chloride contaminated concrete mix (FSEL 1999). Equations 5.3 and 5.4 calculate the percentage of
chlorides per weight of concrete, and percentage of chlorides per weight of cement respectively.
[(9.6lb/yd3)(0.607) / (4258lb/yd3 + 9.6 lb/yd3)] x 100% = 0.137%

[5.3]

[(9.6 lb/yd3)(0.607) / 393 lb/yd3] x 100% = 1.48%

[5.4]

The percentage of chlorides in the chloride contaminated mix well exceeds the threshold of 0.4%
chlorides by weight of cement for cast-in-chlorides. A survey by the Building Research Establishment has
found that the possibility of corrosion is highly probable if the percentage of chlorides by weight of
cement exceeds 1% (Allen, et al. 1993).
5.2.3 Simulation of Existing Field Conditions
In order to simulate existing cracks within a structure, some of the specimens were subjected to flexural
cracking prior to being wrapped. The cracks were created by applying a point load in the center of the
specimen using a Universal Testing Machine. The specimens were loaded until crack widths of 0.01 to
0.013 in. were observed, which is the maximum crack width allowed for exterior exposure by ACI 10.6.4
(ACI 1995). The cylinders were loaded on two sides; the rectangular blocks were only loaded on the
topside. Figure 5.5 shows a typical cracked specimen.

Figure 5.5 Cracked cylinder specimen

30

Two types of repair material, an epoxy grout and a latex-modified concrete, were chosen. Both repair
materials are approved for use on TxDOT projects. The materials were manufactured by Sika
Corporation. The epoxy grout used was Sikadur 42, Grout Pak, and the latex-modified concrete was
SikaTop 122 Plus.
Portions of concrete on the specimens were removed to the level of the reinforcing with a chipping
hammer. All loose dust was removed with a pressure air hose, and the reinforcing was scrubbed with a
wire brush to remove any dirt and/or corrosion product. The repair material was applied according to the
manufacturers specifications. Formwork was used for large patches. Small repair areas were dry packed
(Fuentes 1999).
A surface applied corrosion inhibitor was applied to some of the specimens prior to the application of the
wrap. Sika Ferrogard 903, manufactured by Sika Corporation was applied according to the
manufacturers specification. It was applied with paint rollers to the concrete surface after it was cleaned
with a pressure air hose (Fuentes 1999). Ferrogard 903 is a modified amino alcohol inhibitor that migrates
to the steel reinforcing in order to form a protective coating on the steel surface (Sika 1996).
5.2.4 Wrapping Materials and Procedures
The specimens were either wrapped with Fibrwrap, or a similar generic brand, or were not wrapped. The
wrap was applied according to the manufacturers instructions.
The fabric was saturated with resin by paint rollers. Three layers of FRP wrap were applied by the hand
layup method with a 6 in. overlap. After each layer was applied, air bubbles were pushed out by hand
before the next layer was applied. After the composite had achieved a tacky feel to the surface, the
specimens were examined for any voids. A thickened epoxy mix was injected into the voids. The
composite was painted with Sherwin Williams Hi Bald Aliphatic Polyurethane within 72 hours of
application in order to provide UV protection (Fuentes 1999).
Fibrwrap refers to the TYFO S Fibrwrapsystem manufactured by Hexcel Fyfe Co. This system consists
of TYFO SEH 51 saturated with TYFO S epoxy matrix. The TYFO S Fibrwrap System is the most
common GFRP composite wrapping system used in civil infrastructure. This system was used in the
studies in Canada, California, and Indiana that were discussed in Chapters 1 and 3.
TYFO S Fibrwrap System has been tested extensively to demonstrate its ability to increase strength and
ductility of concrete without increasing the stiffness. The system is also designed to expand when
corrosion causes expansion of concrete. The system has undergone 1000 hour testing for ozone, 140F
temperatures, -40F temperatures, water, salt water, alkaline soil, and ultraviolet light. The testing showed
no significant loss in strength and no failure modes due to environmental effects (Fyfe 1995).
TYFO SEH 51 is a woven fabric of E-glass rovings, with a weight of 27.2 oz/yd2. In the 90, or weft,
direction of the fabric, Kevlar fibers are woven in with the glass fibers to increase strength in the vertical
direction. TYFO S epoxy system is a two-part ambient temperature epoxy resin matrix. (Delta 2000).
The generic system was developed by the IMPACT Laboratory of the Texas Materials Institute. The
generic system consisted of Knytex Reinforcement Fabric A 260-50 by Owens Corning, which is a
unidirectional, woven E-glass fabric. It has a weight of 25.7 oz/yd. Two resin systems were used. One
was an epoxy system similar to the Fibrwrap system, and the other was a vinyl ester resin. The epoxy
system was manufactured by Shell Chemical Company and consisted of EPON Resin 862 with a
polyamine curing agent, EPI-CURE 3234. The vinyl ester system was DERAKANE 411-C50
manufactured by the Dow Chemical Company (Fuentes 1999). The generic systems ability to withstand
environmental effects is currently being tested.
Some specimens were not wrapped in order to serve as controls. This was to allow for evaluation of
single parameters, and to compare the overall performance of the composite material to specimens that
were not repaired with the FRP wrap.
31

Since the length of the wrap may vary on bridge columns, two different lengths of FRP wrap were used
on the cylinder specimens. The first length was 24 in., which simulated wrapping to the waterline. The
wrap was placed on the upper two feet of the specimens, leaving the bottom one foot of the specimen
unwrapped. Therefore the portion of the column that would be subjected the most to the wet/dry cycles
was not wrapped. The second length was 36 in., which simulated wrapping the entire length of the
column. In this case, a larger portion of the composite would be subjected to the wet/dry cycles. The
bottoms of the cylinder specimens were not wrapped since it is not feasible to wrap the bottom of a
column.
For the rectangular blocks, the downstream end was wrapped along with all four sides of the block. The
upstream end, which has the reinforcing extending out, was not wrapped. The wrap length for the
rectangular blocks varies from 24 to 36 in.
The last parameter was the surface condition at the time of the application of the wrap. Since columns in a
marine environment may require wrapping below the waterline, the composite material has to adhere to
the wet surface. Four cylinder specimens were placed in buckets containing 3.5% saline water for
24 hours prior to encapsulation. The remaining twenty-five specimens had air-dried surfaces at the time of
encapsulation (Fuentes 1999).
Access holes, 1 -in. in diameter, were cored into the FRP composite in order to provide access to the
concrete surface for the reference electrode. The holes are located at a distance of 16 in. from the bottom
of the cylinder specimens, which is 4 in. above the waterline. On the rectangular blocks, the access hole is
located on the lower left corner of the downstream end. The access holes are sealed during exposure in
order to prevent moisture from infiltrating. Grease was applied to the exposed steel reinforcing in order to
protect it from corrosion. Figures 5.6 and 5.7 are typical wrapped specimens. Each specimen is described
in detail in Appendix A.

Figure 5.6 Wrapped cylinder specimen

32

Figure 5.7 Wrapped rectangular block specimen

5.3

FIELD PHASE

Field monitoring was conducted on the substructures of highway overpass bridges in Lubbock and
Slayton, TX. Table 5.2 lists the location of the structures in the study. The structures were evaluated
before and after they were repaired and wrapped with FRP composites. The system used was TYFO S
Fibrwrap, which is the same system that is being evaluated in the laboratory phase. The FRP composite
wrap was applied by Delta Structural Technology, Inc. The work was completed in fall 1999.

Table 5.2 Location of structures repaired with FRP composite

Structure #

City

Interchange

#1 - #2

Lubbock

State Loop 289 over Municipal Drive

#3 - #8

Lubbock

US 62/82 & State Loop 289

#9 - #10

Slaton

US 84 over FM 41

#11 - #12

Slaton

US 84 over FM 400

The bridges were experiencing severe damage due to corrosion. The damage consisted of cracking,
spalling, and delamination of the concrete cover on the downstream portions of the bents, and also on
some of the columns. Figures 5.8 to 5.10 show typical corrosion damage found on the substructure. The
damage correlated with the drainage paths for water on the bridge deck. The bents had a slight slope so
that water ran to the lower end as shown in Figs. 5.8 and 5.9.

33

Figure 5.8 Damage to an endcap of Structure #5

Figure 5.9 Delamination on an endcap of Structure #8

34

Figure 5.10 Cracking on column of Structure #12

The project specified the removal of all unsound concrete, and cleaning and/or replacement of
reinforcement in order to remove all of the corrosion products. The concrete was then patched with
Shotpatch 21F by Master Builders Technologies, Inc.
After the damaged areas were repaired, the concrete was sprayed with Sherwin Williams Macropoxy
920 Pre-Prime (Verhulst 1999). Macropoxy 920 Pre-Prime is a rust penetrating epoxy pre-primer
designed for use over marginally prepared surfaces. It may be used as a high performance primer/sealer
(Sherwin Williams 2001).
After the primer was applied, the concrete surface was ground to provide a smooth finish. Then it was
coated with a layer of epoxy thickened with Cab-O-Sil TS 720 manufactured by Cabot Corporation.
The glass fabric was saturated with resin by a saturation machine and then applied to the substructure by
the hand layup method. Three layers of FRP wrap were applied in a manner similar to that described in
the laboratory phase. Figure 5.11 shows Structure #1 after repairs have been made, and before the FRP
composite has been installed. Figure 5.12 shows the parallel structure, Structure #2 after encapsulation.

35

Figure 5.11 Structure #1 prepared for encapsulation

Figure 5.12 Structure #2 encapsulated with FRP wrap

In order to monitor the performance of the FRP composite in a corrosive environment, a non-destructive
method was required. Embedded probes were installed for long-term monitoring, and because the
concrete surface and steel reinforcing were not readily accessible after the structure was wrapped. The
probes that were used were manufactured by Concorr, Inc. The probe consists of a reference electrode,
and a counter electrode encased in a mortar block. The overall dimensions are 2 3/8-in. x 2 3/8-in. x 5 in.
Figure 5.13 and 5.14 are a schematic and cross section of the probe respectively.

36

Figure 5.13 Concorr corrosion rate probe and connection cable (Concorr 1998)

Figure 5.14 Lengthwise section of the corrosion rate probe (Concorr 1998)
The reference electrode is modified graphite, and the counter electrode is a titanium ribbon that is
mounted on the reference electrode. The electrodes are connected to a cable that leads out of the probe. At
the end of the cable is a six-pin connector that connects to the PR-Monitor. The PR-Monitor,
manufactured by Cortest Instrument Systems, Inc., is a corrosion rate measurement device designed for
testing of reinforced concrete. It uses the polarization resistance technique to directly determine the
corrosion rate of the steel reinforcing.
The embedded probes were installed after the damaged concrete was removed and before the repair
material was placed. The probes were placed next to the longitudinal reinforcing. An electrical connection
was made from the probe to the reinforcing. Nine probes were embedded in three separate structures. The
probes were installed near the downstream endcaps of the bent. Figure 5.15 shows a probe after it has
been installed.
37

Figure 5.15 Installed probe on Structure #7

The area of steel being polarized is the surface area of the reinforcing in the direct vicinity of the probe. In
this case, it is the circumference of each reinforcing bar multiplied by the length of the probe. Table 5.3
indicates locations of each embedded probe. The designation left or right is from the perspective of
looking at the structure from the downstream end of the bent.
Table 5.3 Probe installation locations
ID#

Structure

Bent

Beam Face

Distance from End

(ft.)

Steel Area
(in2)

7.1

#7

Left

7.5

44.30

7.2

#7

Right

44.30

8.1

#8

Left

44.30

8.2

#8

Right

44.30

8.3

#8

Left

44.30

8.4

#8

Right

22.15

12.1

#12

Left

44.30

12.2

#12

Right

44.30

12.3

#12

Left

44.30

38

CHAPTER 6: LABORATORY DATA

6.1

SCOPE OF PROJECT

The laboratory specimens described in Chapter 5 were monitored from spring 1999 to spring 2001.
During that time, the specimens were exposed to wet/dry cycles, in order to accelerate the corrosion
process. The wet/dry cycle consisted of a soaking (wet) period of one week in a 3.5% saline solution
followed by a two-week drying period. During the wet period, the lower one foot of the cylinders
(columns) was immersed in salt water. During the drying cycle, the water was removed to a level below
the bottom of the cylinder specimens. This was to create a splash zone effect for the cylinder specimens.
The rectangular block specimens (bents) had saline water irrigated over the top surface. Mats were placed
over the tops to provide even distribution of the water. The blocks were placed at a slight incline,
allowing for the water runoff to flow towards the downstream end. Figure 6.1 is a photograph of the
cylinder specimens in the exposure tank.

Figure 6.1 Cylinder specimens in the exposure tank

Due to the planned long-term exposure studies, this report will concentrate on fourteen specimens that
were monitored, removed from exposure testing, and evaluated. In May 2000, eight specimens, which
will be referred to as Group A, were removed after being exposed to fifteen cycles. In February 2001, six
more specimens, which will be referred to as Group B, were removed after being exposed to twenty-six
cycles. The properties of the specimens are listed in Table 6.1.

6.2

CORROSION MONITORING

Corrosion monitoring is necessary in order to provide some insight into the likelihood of corrosion
activity taking place. This is especially crucial in structures wrapped with FRP composites, since the wrap
prevents visual inspection for signs of corrosion activity, e.g. cracking and staining.
The laboratory specimens were monitored by half-cell potential using a saturated calomel reference
electrode. The procedure followed ASTM Standard C 876, which is described in Chapter 4. The readings
were then converted to equivalent copper/copper sulfate results. The half-cell readings are listed in
Appendix B.

39

Table 6.1 Specimen Parameters

Group A
Specimen*

Cast-inchlorides

FRP Wrap
Fabric
type

Length,
in
24

Resin
system

Initial
concrete
condition

TYFO S

Cracked

Concrete
repair
material+

Corrosion
Inhibitor

CC7

Chlorides

Fibrwrap

CC18

Chlorides

None

Cracked

CNC8

None

Cracked

Ferrogard

CNC13

Generic

24

Epoxy

Cracked

Ferrogard

CNC14

Generic

36

Epoxy

Cracked

Ferrogard

CNC19

Generic

24

Epoxy

Uncracked

RC4

Chlorides

None

RC7

Chlorides

Generic

Cracked
30

Epoxy

Cracked

Group B
CC3

Chlorides

Fibrwrap

24

TYFO S

Uncracked

EG

CC5

Chlorides

Generic

36

Epoxy

Cracked

EG

CC6

Chlorides

Generic

36

Vinyl
Ester

Cracked

EG

CNC10

Fibrwrap

24

TYFO S

Cracked

RNC6

Fibrwrap

30

Epoxy

Cracked

RNC7

None

Ferrogard

LMC

Cracked

* The first letter represents the specimen geometry (C for cylinder, R for rectangular block). The
following letters denote which concrete mix was used (C for cast-in-chlorides, NC for no
chlorides).
+
The notation EG is for the epoxy grout, and the notation LMC is for the latex-modified concrete.

Readings were taken after every four wet/dry cycles for the wrapped specimens during the exposure
period. Readings were taken after every cycle for the unwrapped specimens for the first year and a half of
monitoring. After eighteen months of exposure, it was well established that corrosion was taking place in
the unwrapped specimens. The readings indicated a high probability that corrosion was taking place. In
addition, cracks were forming parallel to the reinforcing in specimens that were not cracked prior to
exposure, as illustrated in Figure 6.2. After eighteen months of exposure, readings were taken after every
four cycles on all of the specimens. Linear polarization testing was performed on the cylinders in Group B
in order to compare results with the amount of corrosion activity found in the specimen after it was
opened.

40

Figure 6.2 Longitudinal cracks forming on an uncracked specimen

6.2.1 Half-Cell Potential Results
The changes in half-cell potential readings over time are illustrated for each group in Figure 6.3 and 6.4.
The half-cell potential readings were generally below 500 mV for the unwrapped specimens, which
indicates a strong likelihood of cracks forming in the concrete due to the formation of corrosion products.
Both cracks and rust stains were observed on the unwrapped specimens. The readings for the wrapped
specimens were typically less negative than the unwrapped specimens. Shortly after the exposure started,
the readings indicated that all of the specimens crossed out of the range of 90% probability of the steel
being passive. The readings fluctuated over time, which is probably due to changes in moisture levels in
the concrete.

Half-Cell Potential
(mV)

Group A
0

CNC8

-100

CC18

-200

CNC14
CC7

-300

CNC19

-400

CNC13

-500

RC4

-600

RC7

-700

90% probability
of corrosion

-800
0

10

20

30

40

50

60

Weeks

Figure 6.3 Half-cell potential vs. time for Group A

41

severe corrosion

Half-Cell Potential
(mV)

Group B
0

CC3

-100

CC5

-200

CC6

-300

CNC10

-400

RNC7

-500

RNC6

-600
-700

90% probability
of corrosion

-800

severe corrosion

20

40

60

80

100

Weeks

Figure 6.4 Half-cell potential vs. time for Group B

The next set of graphs show the readings over time for selected parameters. In Figure 6.5, half-cell
potential for cylinders that were cast with the chloride free mix and have different wrap lengths and
cracking conditions are plotted. Figure 6.6 shows the same parameters for cylinders that had the cast-inchlorides. In Figure 6.7, values for wrapped and unwrapped rectangular blocks for both concrete mixes
are compared.
Uncontaminated Cylinders

CNC8: Unwrapped,
cracked

CNC13: Partial
wrap, cracked

Half-Cell Potential
(mV)

-100
-200

CNC14: Full wrap,

cracked

-300

CNC19, Partial
wrap, uncracked

-400
-500
-600

90% probability of
corrosion

-700

severe corrosion

-800
0

10

20

30

40

50

60

Weeks

Figure 6.5 Uncontaminated concrete cylinders with different parameters

42

Chloride Contaminated Cylinders

CC3: Partial wrap,
uncracked

Half-Cell Potential
(mV)

-100

CC5: Full wrap,

cracked

-200
-300

-500

CC6: Full wrap,

cracked, corrosion
inhibitor
90% probability of
corrosion

-600

severe corrosion

-400

-700
0

20

40

60

80

100

Weeks

Figure 6.6 Chloride contaminated concrete cylinders with different parameters

Rectangular Blocks
0

RNC7: Unwrapped

Half-Cell Potential
(mV)

-100
RNC6: Wrapped

-200
-300

RC4: Unwrapped

-400

RC7: Wrapped

-500
-600

90% probability of
corrosion

-700

severe corrosion

-800
0

20

40

60

80

100

Weeks

Figure 6.7

Half-cell potential vs. time for rectangular blocks

For the cylinders without the cast-in-chlorides, there was not a noticeable difference in the half-cell
potential values for the different parameters. The values suggested that corrosion activity was most likely
taking place in all of the specimens. For the cylinders with the cast-in-chlorides, there was a noticeable
difference between the values for the uncracked specimen compared to the values for the cracked
specimens. For the rectangular blocks, the wrapped specimens had less negative values than the
unwrapped specimens. The readings from the unwrapped specimens were very close in value. For the
wrapped specimens, the specimen with the cast-in-chlorides was more negative than the non-chloride
specimen.

43

The variation in readings for the different parameters is likely to be partially due to the location where the
readings were taken. For the rectangular blocks, the readings were taken at the downstream end. The
downstream end for the unwrapped specimens was subjected to the saline water runoff, whereas the
concrete at the downstream end for the wrapped specimens was much more likely to stay dry because it
was covered with the wrap.
Due to the nature of the wrapping process, the partially wrapped cylinder specimens had the same amount
of unwrapped concrete in the water as the unwrapped specimens. Since the half-cell potential
measurements for the cylinders was taken 4 in. above the waterline, it is likely that the measurements for
the partially wrapped specimens were affected by corrosion activity that might be taking place in the
unwrapped portion below. Also, unlike the rectangular block, moisture could infiltrate the concrete above
the splash zone by capillary action. Therefore the section of the wrapped cylinder specimens that were
being monitored were not as likely to remain as dry as the downstream ends of the rectangular blocks.
6.2.2 Linear Polarization Results
Linear polarization was used on the cylinder specimens in Group B in order to determine the corrosion
rate of the steel reinforcing. The test was performed at two locations on each specimen. The first location
was at 4 in. above the waterline, which allowed for comparison with the half-cell potential readings, see
Table 6.2. The second location was in the middle of the splash zone, 6 in. above the bottom of the
specimen, which is where the most corrosion activity was expected to occur.
The PR-Monitor uses a copper/copper sulfate reference electrode with a sensor controlled guard ring. The
circuit is completed with a large clamp that is used to make an electrical connection to the reinforcing
steel. There is a five-pin connector that connects the half-cell and the sensor controlled guard ring to the
instrument. Prior to running the test, the instrument monitors the free corrosion potential for a period of at
least two minutes to confirm that there is no potential drift present that might affect the accuracy of the
measurements. After the drift is within acceptable limits, less than 2 mV/min, the instrument starts
measuring the overvoltages. At the end of the polarization cycle, an AC signal is applied from a high
frequency generator in order to measure the solution resistance of the concrete. The computer then
calculates the polarization resistance and the corrosion rate. Figure 6.8 shows the PR-Monitor in place for
the linear polarization testing of the cylinder specimens.

Table 6.2 Comparison of Half-Cell Potential and PR-Monitor

Specimen

Ecorr Half-Cell Potential

(mV)

Ecorr PR-Monitor
(mV)

CC3

-212

-275

CC5

-460

-411

CC6

-459

-458

CNC10

-356

-275

44

Figure 6.8 PR-Monitor test setup on specimen CC5

The testing was performed after the FRP composite wrap had been removed from the specimens. The
clamp was connected to the reinforcing bar that was being polarized. The test could not be performed on
the lower portion of specimen CC3 because the epoxy resin from the composite had encased the entire
portion of concrete below the wrap. The results from the tests are shown in Table 6.3 and Figure 6.9.
The values for Ecorr from the half-cell potential testing and the linear polarization testing agree well. The
first set of values was taken when the specimens were removed from the exposure tank. They were taken
with the saturated calomel reference electrode and then converted to the equivalent copper/copper sulfate
reading. The second set of values was obtained from the linear polarization testing, which was six weeks
later. They were taken with a copper/copper sulfate reference electrode. The Ecorr measurements were
consistently higher in the splash zone compared to those above the splash zone, as noted in Table 6.3.
The results from the linear polarization tests indicate that severe corrosion is taking place in most of the
specimens. Specimen CC3 was the only specimen to fall into the moderate corrosion rate level. From
visual observations of the specimens, there were no outward signs of corrosion activity on the middle
portions of CC3 and CNC10. There was minimum cracking on CC3, and CNC10 did not have the cast-inchlorides. The lower portion of CNC10 showed signs of corrosion activity as seen in Figure 6.10. Rust
stains were visible at both locations on specimen CC6, and in the splash zone of CC5.
It should be pointed out that in this report, the corrosion current density is denoted as icorr as well as
Icorr. In most textbooks, the symbol for corrosion current density is icorr and the symbol for corrosion
current is Icorr. However, some manufacturers use the symbol Icorr for corrosion current density in their
literature; hence it is given both ways in this report.

45

Table 6.3 Linear Polarization Data from Group B

Location

Specimen

Splash zone

4 in. above the waterline

N/A

-248

Icorr (A/cm )

N/A

0.74

Rate (mpy)

N/A

0.34

-475

-411

Icorr (A/cm )

4.88

4.27

Rate (mpy)

2.23

1.95

-504

-458

3.09

5.23

1.41

2.39

-549

-275

Icorr (A/cm )

8.45

1.27

Rate (mpy)

3.86

0.58

CC3
Ecorr (mV)
2

CC5
Ecorr (mV)
2

CC6
Ecorr (mV)
2

Icorr (A/cm )
Rate (mpy)
CNC10
Ecorr (mV)
2

High Corrosion

Specimen

CNC10

Middle of specimen
Splash zone

CC6

CC5

CC3
0

10

Corrosion Current Density

Figure 6.9 Corrosion current density (A/cm2) values for Group B

46

Figure 6.10 Evidence of corrosion activity in the splash zone of specimen CNC10

6.3

TESTING OF SPECIMENS

Before opening the specimens, they were allowed to dry in order to remove any moisture remaining in the
concrete. After Group A was removed from the exposure tank; it was left to air dry for six weeks prior to
being evaluated. Group B was left to air dry for eight weeks prior to being evaluated. The cracks on the
specimens were mapped and measured to determine the largest crack size observable. Chloride content
determination tests were run on all of the specimens. Pull-off bond tests were run on the wrapped
rectangular block in Group B.
6.3.1 Crack Mapping
Since cracks provide direct access for oxygen, moisture, and salts to enter the concrete, it is important to
note their size and location. ACI recommends that cracks should be smaller than 0.004 in. in order to be
watertight (ACI 1992). Group A was mapped prior to the wrap being removed. Group B was mapped
after the wrap was removed. The crack locations were marked on the concrete. Then the crack size was
measured with a crack comparator. Table 6.4 lists the maximum crack sizes observed.
It is likely that there were larger cracks on most of the wrapped specimens in Group A that were not
observed due to being covered by the FRP wrap. The two uncracked specimens, CNC19 and CC3, both
had hairline cracks that formed during the exposure period. The rest of the specimens had cracks larger
than the recommended 0.004 in.

47

Table 6.4 Crack Size Data

Specimen Maximum Crack Size (in)
CC7
CC18

0.01
0.02

CNC8

0.007

CNC13

0.016

CNC14

N/A: Fully wrapped

CNC19

0.002

RC7

N/A: Fully wrapped

RC4

0.025

CC3

0.002

CC5

0.016

CC6

0.009

CNC10

0.005

RNC6

0.013

RNC7

0.025

6.3.2 Chloride Determination

In addition to crack size, it is important to determine the chloride content of the concrete. The chloride
content indicates whether the chloride threshold has been reached, and whether the FRP composite wrap
prevents the ingress of chlorides. The chloride threshold is approximately 0.03% chloride by weight of
concrete. The chloride content calculated for the cast-in-chlorides mix was 0.137% chloride by weight of
concrete.
Chloride content tests were run on the specimens to determine the chloride content in the concrete after
exposure. A James Instruments CL-500 test was used. Samples were taken using a hammer drill with a
-in. bit. For Group A, the first -in. of concrete was removed and discarded, and then samples were
taken to the depth of 1 -in. For the cylinders, the sample locations were 6 in. from the top, and 6 in.
from the bottom. For the rectangular blocks, the sample locations were the center of the top face of the
block, and the center of the side face of the block. For Group B, the chloride content tests were used to
determine if the chloride levels changed at different depths of the cover. The first -in. was discarded,
and then samples were taken at depths of -in. and 1 in. The concrete cover over the reinforcing steel was
located at the depth of 1 in. The samples were taken at the same level that the half-cell potential
measurements were taken. This was at 4 in. above the water line for the cylinders, and at the lower left
corner of the downstream end for the rectangular blocks. The results from the chloride content tests are
listed in Tables 6.5 and 6.6.

48

Table 6.5 Results from chloride tests for Group A at 1 -in. depth
Specimen
Cylinders

Chloride Content by Weight of Concrete

6 in. from top

6 in. from bottom

% Cl

Surface

% Cl-

Surface

CC7

.14

FRP

.30

Bare

CC18

.62

Bare

.26

Bare

CNC8

.001

Bare

.33

Bare

CNC13

.002

FRP

.24

Bare

CNC14

.002

FRP

.04

FRP

CNC19

.002

FRP

.26

Bare

Rectangular Blocks

Top

Vertical

% Cl-

Surface

% Cl-

Surface

RC4

.33

Bare

.36

Bare

RC7

.12

FRP

.12

FRP

Table 6.6 Results from chloride tests for Group B

Specimen
Cylinders

Chloride Content by Weight of Concrete

Depth of -in.

Depth of 1 in.

% Cl

Surface

% Cl-

Surface

CC3

.09

FRP

.08

FRP

CC5

.16

FRP

.08

FRP

CC6

.16

FRP

.13

FRP

CNC10

.003

FRP

.003

FRP

Rectangular Blocks

Depth of in.

Depth of 1in.

% Cl-

Surface

% Cl-

Surface

RNC6

.002

FRP

.002

FRP

RNC7

.21

Bare

.21

Bare

The difference in levels of chlorides found in the rectangular block specimens indicates that the FRP wrap
provides a barrier for chlorides. This is evident by the very low levels of chlorides found in RNC6 as
compared to RNC7. Both specimens were cast with the chloride-free mix and exposed in the same way.
The percentages of chlorides in the unwrapped specimen were two orders of magnitude larger than the
wrapped specimen. In specimens RC4 and RC7, the chloride levels in the wrapped specimen were
slightly lower than the predicted chloride levels for the chloride contaminated specimens. The chloride
levels in the unwrapped specimen were considerably higher.

49

The samples from the lower portion of the cylinders, which was located in the splash zone, had a chloride
content of 0.24 to 0.33% for all exposed concrete. Specimen CNC14, which was a chloride free specimen,
was the only cylinder that was wrapped the full length. It had the lowest chloride content in the splash
zone for the cylinders. The chloride content in the splash zone of CNC14 was still above the threshold of
0.03%. This indicates that the chlorides permeated through the bottom of the column in order to
contaminate the concrete.
The specimens in Group B had very little difference in chloride levels with respect to depth of the
concrete cover. The variance in chloride levels from 0.08 to 0.16% for the wrapped chloride contaminated
specimens was most likely due to some unevenness in the distribution of the chlorides in the mix and that
some of the powder samples may have contained portions of the larger aggregates, which would not
contain chlorides.
6.3.3 Bond Test
The strengthening benefits of FRP composite wrapping systems are dependant on a strong adhesion
between the wrap and the concrete surface. Without sufficient bond, the external strengthening benefits of
the wrap are lost. When FRP composite wrap is applied by the hand layup method, the resin also serves as
the system adhesive. If this adhesive layer deteriorates due to environmental effects, the bond between the
concrete surface and the FRP composite deteriorates.
The bond of the FRP composite wrap was tested on specimen RNC6. The rectangular block specimen
was chosen because the geometry was not as conducive to applying the wrap as the cylinders. It was
difficult to maintain the tension required to apply the wrap (Fuentes 1999). Also a flat surface was
required in order to perform the pull-off test.
The test was performed with a DYNA Z16 Pull-off Tester manufactured by Proceq. The pull-off tester
measures the load required to cause a tensile failure in the concrete. From it, the compressive strength of
the concrete can be estimated. The bond between the concrete and repair material can be evaluated by the
location of the failure (Long and Murray 1995).
To carry out the test, two 2 in. diameter cores were partially cored into the center of the top face and the
side face of the specimen to a depth of -in. One was near the upstream end and one was near the
downstream end. A 50 mm. steel disk was then epoxied onto the FRP composite wrap at the end of the
core. The disk was then fastened into the DYNA Z16 Pull-off tester, and slowly pulled until failure
occurred. The device measured the force at failure. Figure 6.11 shows the test setup and Table 6.7 lists the
results from the pull-off tests.

Figure 6.11 Test setup for DYNA Z16 Pull-Off Tester

50

Table 6.7 Results from pull-off test for RNC6

Depth of
Failure

Failure
Strength

Failure
Strength

(in)

(N/mm2)

(psi)

Top, downstream end

0.7

101.5

Top, upstream end

1.22

176.9

Side, downstream end

1/8

1.3

188.5

Side, upstream end

1/8

1.63

236.4

Test
#

Location

All of the tests failed in the substrate except for test #2, which failed at the bond between the wrap and the
concrete surface, as shown in Figure 6.12. This indicates that the bond strength was stronger than the
concrete tensile strength for most of the specimen. The top face had the most severe exposure, so it is
reasonable that the tensile strength was lower. The upstream end was most likely to have moisture seep
under the wrap, which may have resulted in the failure at the epoxy bond in that location. The differences
in failure strength are also possibly due to the fact that concrete is not an isotropic material. In addition,
the low failure strength in the substrate during test #1 may have also been caused by cracks in the
specimens. The difference in values of failure stress does indicate that there is some variation in the
adhesion of the FRP wrap to the concrete surface.

Figure 6.12 Failure at the epoxy/concrete interface on Test #2

The pull-off strengths agreed well with the study by Abu-Tair, Burley, and Rigden of bond strength of
repair materials subjected to different loading and exposure conditions. In their pull-off tests, the average
failure stress for concrete repaired with an epoxy mortar, a polymer modified cementitious mortar, and a
portland cement concrete repair material was 1.63 N/mm2, 1.39 N/mm2, and 1.06 N/mm2 respectively
(Abu-Tair, Burley, and Rigden 1995).

6.4

OPENING OF SPECIMENS, GROUP A

The unwrapped specimens were opened by first removing the outer 1 in. of concrete cover with a
chipping hammer. The concrete core was then examined for corrosion products. Then the concrete core

51

was broken into small pieces with the chipping hammer. The concrete pieces were removed, and the
remaining steel reinforcing cage was then examined for signs of corrosion activity.
The FRP composite wrap prevented easy removal of the concrete cover. The specimens were scored four
times lengthwise with a concrete saw in order to provide access to the concrete surface. The concrete
cover along the score was loosened with the chipping hammer. The wrap generally pulled away from the
concrete along the score. This allowed for the wrap to be removed. Although the concrete along the
scores was damaged due to the chipping hammer, there was very little damage to the rest of the concrete
surface after the wrap was removed. Once the FRP composite wrap was removed, the specimens were
opened and examined the same way as the unwrapped specimens. Table 6.8 summarizes the observations
noted for Group A.
Table 6.8 Observations from opening of Group A
Specimen

Corrosion activity

CC7

The top 2 were wrapped. Epoxy that had dripped down from the wrap covered half of the exposed
surface on the bottom 1. Heavy corrosion was found on bars and stirrups near cracks. Minor
corrosion was found throughout the upper 2.

CC18

The entire surface was exposed. A small honeycombed area was located at the bottom of the
specimen. Areas of heavy corrosion were found throughout on the bars and stirrups near cracks and
also near the honeycombed area. A small amount of section loss was found at the lower ends of
two bars near the honeycombing.

CNC8

The entire surface was exposed. A honeycombed area was at the bottom of the specimen. Heavy
corrosion with substantial section loss was found on the bottom stirrup near the honeycombing.
Moderate corrosion was found on the lower half of the bar by the honeycombing and on the lower
5 stirrups.

CNC13

The top 2 were wrapped. A corrosion inhibitor was applied. Light to moderate corrosion was
found on the lower 6 of the bars and on the bottom 2 stirrups. Corrosion activity corresponded
with crack locations.

CNC14

The entire surface was wrapped. A corrosion inhibitor was applied. Moisture was found trapped
beneath the wrap in the splash zone. Minor to moderate corrosion was found on the lower 6 of 1
bar and on the bottom 2 stirrups near where the moisture was found.

CNC19

The top 2 were wrapped. The specimen was not cracked prior to exposure. Moderate corrosion
was found near crack locations on the lower 1 of 2 bars. Minor corrosion was found on the lower
3 stirrups.

RC4

The entire surface was exposed. Moderate corrosion was found throughout the bottom 2 bars and
on all 3 stirrups.

RC7

The entire surface was wrapped. Moisture was found trapped underneath the wrap on the bottom
face at the upstream end. Moderate corrosion was found on the upper stirrup near where the
moisture was found. Minor corrosion was found throughout the bottom 2 bars.

Corrosion activity generally corresponded with cracks and honeycombing found within the concrete.
Specimens with cast-in-chlorides had corrosion activity throughout the specimen whereas the specimens
without cast-in-chlorides had corrosion only in the splash zone and a few inches above the splash zone.
The wrapped specimens had less corrosion activity than the unwrapped specimens. Moisture was found
trapped between the FRP composite wrap and the concrete surface in specimens CNC14 and RC7, both of
which were fully wrapped. Figure 6.13 shows moisture that was found at the interface of the wrap and the
concrete surface on specimen CNC14.

52

Corrosion activity appeared less severe on the specimens that were treated with the Sika Ferrogard 903.
Corrosion activity in specimen CNC19, which was not treated with the corrosion inhibitor, was found on
the stirrups throughout the splash zone even though the specimen was not cracked prior to exposure.
Corrosion activity in specimens CNC13 and CNC14, which were treated with the corrosion inhibitor, was
near the crack locations. Very little corrosion activity was found in the uncracked portions of CNC13.

Figure 6.13 Moisture trapped beneath the surface in the splash zone of specimen CNC14

6.5

OPENING OF SPECIMENS, GROUP B

In order to perform the linear polarization testing, the FRP composite wrap had to be removed with
minimal damage to the concrete surface. It was noticed in Group A that the wrap could be pried away
from the concrete along the scores. Therefore the specimens were scored twice lengthwise with a concrete
saw. The wrap was then pried off with a crowbar. The epoxy resin that had dripped down from the
composite and encased the bottom portion of CC3 could not be removed.
The generic system with the vinyl ester was the easiest to remove. The Fibrwrap system was the most
difficult to remove. This method resulted in very little damage to the concrete surface. As a result, the
entire surface of the wrapped specimens could be examined for cracks, and the linear polarization testing
could be performed. After the wrap was removed, the specimens were opened the same way as Group A.
Table 6.9 summarizes the observations noted for each specimen for Group B.

53

Table 6.9 Observations from opening of Group B

Specimen

Corrosion activity

CC3

The top 2 were wrapped. The bottom 1 was covered with epoxy resin that had dripped down
from the wrap. The specimen was not cracked prior to exposure. An epoxy grout patch covered
the bottom 1 of the specimen. Corrosion activity was found mainly in the splash zone, which was
also where the patch was located. A small amount of moderate corrosion was found at the lower
ends of the bars. Half of the surface area of the bottom 4 stirrups was covered with minor to
moderate corrosion.

CC5

The entire surface was wrapped. An epoxy grout patch covered the top 6 of the specimen. A
small honeycombed area was located at the bottom of the specimen. Small areas of moderate
corrosion were found on the bars near cracks throughout. Minor corrosion was found on the lower
6 stirrups, with moderate corrosion on the bottom stirrup at the honeycombing. Heavy corrosion
with minor section loss was found on the top stirrup at the patch.

CC6

The entire surface was wrapped. Moisture was trapped beneath the wrap in the splash zone. An
epoxy grout patch covered the top 1 of the specimen. A corrosion inhibitor was applied.
Moderate corrosion was found on bars at crack locations and near where the moisture was found.
Minor corrosion was found on the lower 5 stirrups. Heavy corrosion was found on the upper 4
stirrups at the patch.

CNC10

The top 2 were wrapped. Corrosion activity corresponded with crack locations. Heavy corrosion
was found on the lower 6 of bars and on lower 2 stirrups. One bar had an area of moderate
corrosion just above the splash zone. No corrosion activity was found on the upper 18 of the
specimen.

RNC6

The entire surface was wrapped. A small amount of moisture was trapped at the upstream end of
the top face of the specimen. A latex-modified concrete patch covered the downstream end of one
of the side faces. Isolated minor corrosion was found along the bars and stirrups near the patch.
Heavy rust was found on all of the bars near the end of the wrap. Heavy rust with substantial
section loss was found on the top stirrup.

RNC7

The entire surface was exposed. A longitudinal crack had formed at the location of the bottom
bars. Heavy corrosion was found throughout the bars, with moderate to heavy corrosion found on
stirrups near crack locations.

Due to the longer exposure time, the corrosion activity was more severe in Group B than Group A.
Corrosion activity also corresponded with the crack locations, although more corrosion was found away
from cracks. This is mainly due to the fact that most of the specimens had cast-in-chlorides.
Moisture was found trapped beneath the wrap for two of the fully wrapped specimens, CC6 and RNC6.
No moisture was found in specimen CC5, even though it was fully wrapped. Corrosion activity near the
location of trapped moisture was greater than in other areas of the specimen. Specimen RNC6
experienced section loss in the stirrup due to the trapped moisture as shown in Figure 6.14.
Unlike Group A, corrosion activity was also found in the upper half of the cylinder specimens. Heavy
corrosion was found at the patches in the chloride contaminated concrete indicating that macrocells had
formed between the concrete and the uncontaminated repair material for the cylinder specimens. The
macrocells were found both in the splash zone and several feet above the splash zone in encapsulated
concrete. Figure 6.15 shows the corrosion activity found in the top portion of specimen CC5. A macrocell
also formed between the uncontaminated concrete and the repair material in RNC6, but the corrosion
activity was not as extensive. This was most likely due to fact that the existing concrete did not contain
chlorides.

54

Figure 6.14 Section loss on the upstream stirrup of specimen RNC6

Figure 6.15 Macrocell formation in the top portion of specimen CC5

There was no sign of decreased corrosion due to the use of the corrosion inhibitor in Group B. Corrosion
activity was more extensive throughout specimen CC6, which was treated with Sika Ferrogard 903,
compared to CC5 which was not treated.

55

6.6

DISCUSSION OF RESULTS

Both the half-cell potential readings over time and the linear polarization indicate the likelihood of
corrosion activity taking place. Opening of the specimens found corrosion activity to varying degrees was
taking place in each specimen.
The chloride tests confirmed that the FRP composite wrap prevents chlorides from permeating the
concrete. As a result, very little corrosion activity was found in portions of uncontaminated concrete that
was free of cracks and repair material. However, corrosion activity was present throughout the wrapped
portions of the chloride contaminated specimens. This indicates that although the FRP composite wrap
provides a barrier to chlorides and moisture, corrosion activity may continue in the cracked chloride
contaminated concrete. Corrosion activity was also accelerated by the formation of macrocells between
the existing concrete and repair material.
Corrosion activity was found in the splash zones of all of the cylinder specimens. Chlorides were found in
the lower portions of all of the cylinder specimens including the fully wrapped cylinder without the castin-chlorides. The FRP composite wrap did not prevent chloride ingress in and just above the splash zone.
Lastly, in four out of five of the fully wrapped specimens, moisture was found trapped beneath the wrap
resulting in accelerated corrosion activity in that area. For the cylinder specimens, it resulted in corrosion
activity taking place in the splash zone. For the rectangular blocks, the area of corrosion activity due to
trapped moisture was found at the upstream end rather than the downstream end. As a result, the activity
did not register on the half-cell potential monitoring since the downstream end was the area being
monitored. The results from the linear polarization showed that the half-cell potentials varied depending
on the location where they were taken. The ASTM standard suggests an interval of 4 ft. for a bridge deck.
In the case of the laboratory specimens, the half-cell potential could vary substantially in an interval of
10 in.
The following figures help to quantify the corrosion rate measurements taken from the linear polarization
testing. Figure 6.16 shows the steel reinforcing that was polarized in the splash zone of CNC10. It had the
highest corrosion rate measurement, 3.86 mpy. The concrete was not wrapped with FRP composite in that
area. Figure 6.17 shows the steel reinforcing that was polarized in the splash zone of CC6. The concrete
was wrapped with FRP composite in that area. The corrosion rate was 1.41 mpy. Figure 6.18 shows the
steel reinforcing that was polarized in the middle portion of CC3. It had the lowest corrosion rate
measurement, 0.34 mpy, and the only measurement that was in the moderate corrosion rate level. The
concrete was wrapped with FRP composite in that area. As shown in the following figures, the corrosion
rate was consistent with the extent of corrosion activity. Photographs of the steel reinforcing cage for each
specimen in both groups may be found in Appendix C.

56

Figure 6.16 Steel reinforcing of CNC10 with a corrosion rate of 3.86 mpy

Figure 6.17 Steel reinforcing of CC6 with a corrosion rate of 1.41 mpy

57

Figure 6.18 Steel reinforcing of CC3 with a corrosion rate of 0.34 mpy

58

CHAPTER 7: FIELD DATA

7.1

SCOPE OF PROJECT

The field installation described in Chapter 5 has been monitored for twelve months during 2000 to 2001,
visually and using embedded probes. The condition of the structures was evaluated prior to repair and
encapsulation with FRP composite wrap. Readings from the embedded probes have been monitored
approximately every six months. The objective is to determine the likelihood of corrosion activity
continuing after the FRP composite was installed.

7.2

FIELD RESULTS PRIOR TO REPAIRS

In the summer of 1998, the structures were evaluated to determine the extent of the corrosion activity.
The testing was performed prior to repairs. Three different linear polarization devices were used: the 3LP
device, the PR-Monitor, and the Gecor device. This was done to determine how well each device operated
in the field. The Ecorr values for the PR-Monitor and the Gecor device were more consistent than the 3LP
device. The corrosion current density was higher on the PR-Monitor than the Gecor device. The 3LP
device produced inconsistent results, which may have been due to technical difficulties with the reference
electrode. The linear polarization was performed 10 to 20 feet from the downstream end of the bent in
order to be able to connect to the steel and to avoid the delaminated areas. The results indicated that the
steel was mostly passive. This is reasonable since the measurements were taken at places that showed
little to no damage (Verhulst 1999). The results are summarized in Table 7.1 and Figure 7.1.
The chloride content with respect to depth was determined at different locations. This was done to
classify the possibility of corrosion conditions for the structures. The samples were taken using a hammer
drill with a -in. bit. The first -in. was discarded to eliminate surface imperfections or inconsistencies.
Samples were taken at the depth of 1 in., 1.5 in., and 2 in. The results are listed in Table 7.2.

Table 7.1 Corrosion Rate Measurement Data (July 1998)

Location (Structure #)
Device
#8

#7

#1

#5

#10

#11

-40

N/A

-100

-270

N/A

-157

0.1347

N/A

0.0431

0.063

N/A

0.6207

-108

-79.2

-102.7

-74.9

-66.8

-144.5

0.0219

0.46

0.0044

0.018

0.15

0.302

0.01

0.21

0.002

0.008

0.069

0.138

-83.8

-78.3

-116.4

-41.4

-49.8

-113

Icorr (A/cm )

0.01

0.001

0.009

0.003

0.008

0.022

Interpretation

Passive

Passive

Passive

Passive

Passive

Passive

3LP
Ecorr (mV)
Icorr (A/cm2)
PR-Monitor
Ecorr (mV)
2

Icorr (A/cm )
Rate (mpy)
Gecor
Ecorr (mV)
2

59

Passiv

Low

Moderat

Structure
Structure

Gecor Device
PR-Monitor
3LP Device

Structure
Structure
Structure
Structure
0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Corrosion Current Density (July 1998)

Figure 7.1 Corrosion current density (A/cm2) values from different equipment

Table 7.2 Chloride percentages by weight of concrete (FSEL 1999)

Depth (in.)

Location
0.5 1.0

1.0 1.5

1.5 2.0

Structure #8 10 ft. from the downstream end.

0.12

0.034

0.038

Structure #8 2.5 ft. from the downstream end.

0.17

0.21

0.20

Structure #7 20 ft. from the downstream end.

0.19

0.18

0.15

Structure #7 12 ft. from the downstream end.

0.21

0.28

0.16

Structure #3 22 ft. above the ground on the west face

of the column.

0.26

0.29

0.19

Structure #2 17 ft. from the downstream end.

0.31

0.22

0.16

Structure #2 25 ft. from the downstream end (left of

the center column).

0.056

0.08

0.042

Structure #1 Directly on the downstream endcap.

0.01

0.0056

0.0035

Structure #5 Directly on the spalled downstream

endcap.

0.45

0.38

0.21

Structure #5 10 ft. from the downstream end (some

spalling present).

0.082

0.043

0.0035

Structure #10 Between columns away from the

downstream end.

0.003

0.003

0.003

Structure #11 On the top of the bent.

0.018

0.02

0.018

60

The shaded boxes indicate the three locations where the chloride threshold had not been reached. The
chloride levels in most of the locations were higher than the threshold of 0.03% chloride by weight of
concrete. This means that many of the locations tested have a sufficient level of chlorides to allow for the
onset of corrosion. The highest levels of chlorides were found in areas that were already damaged.
However, high levels of chlorides were found in Structure #7 and #2 in locations that were 10 to 25 feet
away from the downstream end.
A pachometer was used to determine the depth to the steel reinforcing. The depth of cover was found to
be 2.5 in. Given that eight of the twelve locations were above the threshold of 0.03% at a depth of 2 in., it
is likely that chloride contamination is occurring at the depth of the reinforcing steel (Verhulst 1999).

7.3

MONITORING RESULTS

Three trips were made to Lubbock, TX in order to collect data from the field installations. Readings were
taken with the PR-Monitor. The tops of the bridge bents were accessed using a lift truck provided by
TxDOT. The lead cables from the embedded probes protruded out from the FRP wrap on the top of the
bents. The six-pinned connector at the end of the cable was plugged into a connecting cable from the PRMonitor. The linear polarization test was run by using a program on the laptop console on the PRMonitor. The PR-Monitor operates in the same manner as described in Chapter 6, except that the
embedded probe replaces the half-cell and sensor controlled guard ring assembly. Figure 7.2 shows the
PR-Monitor being used in the field.

Figure 7.2 PR-Monitor test setup for Structure #12

The embedded probes were installed in three bridges: Structure #7, Structure #8, and Structure #12.
Corrosion rate measurements for probe 12.3 could not be taken due to invalid data. Possible damage to
the probe or the connection to the steel may have been the cause of the problem. The linear polarization
tests were performed in May 2000, October 2000, and May 2001. The results are summarized in Table
7.3 and Figure 7.3.

61

Table 7.3 Linear Polarization Data

Probe ID #

7.1

7.2

8.1

8.2

8.3

8.4

12.1

12.2

12.3

Ecorr (mV)

Icorr (A/cm2)

Rate (mpy)

May 2000

-244

0.57

0.26

October 2000

-251

0.37

0.17

May 2001

-253

0.37

0.17

May 2000

-477

1.60

0.73

October 2000

-426

1.09

0.50

May 2001

-528

1.82

0.83

May 2000

-329

0.44

0.20

October 2000

-306

0.24

0.11

May 2001

-344

0.44

0.20

May 2000

-362

1.36

0.62

October 2000

-352

0.94

0.43

May 2001

-384

1.42

0.65

May 2000

-377

0.57

0.26

October 2000

-392

0.33

0.15

May 2001

-395

1.23

0.56

May 2000

-351

0.99

0.45

October 2000

-461

0.79

0.36

May 2001

-359

1.31

0.60

May 2000

-305

3.37

1.54

October 2000

-316

0.59

0.27

May 2001

-392

2.19

1.00

May 2000

-384

2.69

1.23

October 2000

-303

2.08

0.95

May 2001

-436

2.45

1.12

May 2000

N/A

N/A

N/A

October 2000

N/A

N/A

N/A

May 2001

N/A

N/A

N/A

Date

62

Low

Moderate

High Corrosion

12.2

Probe ID #

12.1
8.4

Spring 2001
Fall 2000

8.3

Spring 2000

8.2
8.1
7.2
7.1
0

Corrosion Current Density

Figure 7.3 Corrosion current density (A/cm2) for corrosion monitoring

Many of the half-cell potential values are more negative than 350 mV, indicating that corrosion is likely
to be taking place. Many of the corrosion rate readings indicate that moderate to heavy corrosion activity
is taking place. The readings were higher in the spring than in the fall. This is probably due to higher
temperatures and moisture levels in the spring. The corrosion rate measurements for Structure #12 were
noticeably higher than the other structures.
The corrosion rates from the linear polarization testing were substantially higher than the corrosion rates
from linear polarization testing performed before the repairs were made. The difference is most likely due
to the fact that the embedded probes are located 4 to 7 feet from the downstream end in areas where
corrosion activity was noted before the repairs were made, while the pre-repair linear polarization testing
was performed 10 to 20 feet from the downstream end in areas that did not show signs of corrosion
activity.

7.4

OBSERVATION OF FIELD INSTALLATION

The bents that had the embedded probes were examined each time the linear polarization testing was
performed. Moderate debris accumulation was found on top of all of the bents. The FRP composite on the
bridge bents had been cut around the bearing pads. In some places, the wrap was finished with epoxy to
seal it around the bearing pad; in other places it was not. The finishes around the bearing pads near probes
7.1, 8.3, and 8.4 seemed to have the best seals. There was a 1 in. gap between the wrap and the bearing
pad near the location of probes 12.1 and 12.2. This allowed for exposed concrete on the top of the bent as
shown in Figure 7.4. Figure 7.5 shows the finish around the bearing pad near probe 7.2. Figure 7.6 shows
the finish around the bearing pad near probe 8.3, and is an example of how the wrap around the bearing
pad may be sealed off. However, little is known about the viability of the seal if there are movements of
the girder and bearing pad.

63

Girder ends

Exposed concrete
Wrap

Figure 7.4 Finish of FRP wrap around girders of Structure #12

Girder

Edge of wrap
Figure 7.5 Finish of FRP wrap around girders of Structure #7

64

Girder ends

Wrap
Figure 7.6 Finish of FRP wrap around girders of Structure #8
There is some correlation between the finish of the wrap and the corrosion rate measurements. The
readings from probe 7.1, and the readings from Structure #8, were generally lower than the readings from
probes 7.2, 12.1, and 12.2. Cracks were found on the ends of the girders near probes 8.3 and 8.4. Cracks
with rust stains were found on the ends of the girders near probes 7.2, 12.1, and 12.2. The cracks may be
an indication of exposure to runoff containing deicing salts.
In May 2001, all of the structures were visually inspected. On several of the columns, the FRP composite
wrap was damaged as a result of vehicle impact. This damage will undoubtedly influence the performance
of the FRP composite. In the field study in Indiana, researchers noticed that collision damage to the FRP
wrap worsened over time because the damaged areas were more sensitive to the effects of a harsh
environment. Figure 7.7 shows the worst damage noticed on the structures at the Lubbock site.

Figure 7.7 Vehicular collision damage on column of Structure #2

65

At the time of the visual inspection, the bridge decks were being replaced on Structures #1 to #8 because
of corrosion damage. On many of the bridges substructures there were rust stains that had resulted from
corrosion activity that was taking place in the superstructure above. However, on Structure #1, there were
signs of corrosion activity taking place in one of the bents. Rust stains were found at the seam of the FRP
wrap on the bottom of the fourth bent on both sides of the north column. There were no indications that
the stains had resulted from corrosion activity in superstructure. Figure 7.8 is a photograph of the rust
stains.

Figure 7.8 Rust stains on bent of Structure #1

7.5

COMPARISON OF RESULTS FROM THE LABORATORY AND THE FIELD

The linear polarization and half-cell potentials both indicate a strong probability that corrosion activity is
taking place near the probe locations. The staining found on one of the bents indicates that there may also
be corrosion activity in other locations that are not being monitored.
The chloride measurements indicate high levels of chloride in many locations in the structures. These
locations include areas that were then repaired and areas that were most likely not repaired. The
specifications for the repairs called for the removal of all unsound concrete. The chloride content
determination indicates that the structures are likely to contain sound but chloride contaminated concrete.
This means that there may be macrocells forming between the repair material and the existing concrete
similar to that found in the laboratory specimens. It is also likely that chloride contaminated concrete was
encapsulated with FRP composite wrap.
Moisture may be able to enter into the concrete through the exposed concrete surface around the bearing
pads on the bents where the FRP wrap was not sealed. The laboratory specimens showed that moisture
that seeped into the concrete near the edge of the FRP wrap could become trapped, resulting in
66

accelerated corrosion in that area. It is possible that moisture is being trapped in the bents near the bearing
pads. This could be causing accelerated corrosion, which would explain the higher corrosion rate
measurements found in Structure #12.
It should be noted that the embedded probes were originally planned to be installed closer to the endcaps.
This was because the downstream endcaps showed the most damage in the field before repairs were
made. It was expected that the endcaps would be the most susceptible place for any corrosion that might
take place after encapsulation. Due to the construction timing, the embedded probes were installed 4 to 7
feet away from the downstream end of the bent because the endcaps had already been repaired. In most
cases, the embedded probes were installed near the bearing pads.
It is quite possible that the corrosion activity that is taking place near the bearing pads is more severe than
corrosion activity in the endcaps. When the laboratory study was initiated, it was expected that the
downstream ends of the rectangular blocks would be the area that was most likely to experience corrosion
activity. Instead, the most severe corrosion activity was found at the upstream ends of the rectangular
blocks where the moisture had become trapped underneath the wrap.

67

68

CHAPTER 8: CORROSION INHIBITORS

8.1

INTRODUCTION TO CORROSION INHIBITORS

In order to further study the issue of corrosion prevention in FRP encapsulated chloride contaminated
concrete, an additional laboratory study was developed to investigate the effectiveness of surface applied
corrosion inhibitors. Three different commercially available corrosion inhibitors were selected to be
studied.
The objective of this additional study is to find out whether the use of a surface applied corrosion
inhibitor will prevent corrosion activity in chloride contaminated concrete in members where FRP
composite wraps serve as a barrier for moisture and further chloride ingress.

8.2

TYPES OF CORROSION INHIBITORS

Corrosion inhibitors are found in two forms, admixtures and liquid coatings that are applied to the
surface. The first chemicals tested as corrosion inhibitors were additives containing sodium nitrite,
potassium chromate, sodium benzoate, stannous chloride, and calcium nitrite. Calcium nitrite was the
only product that became commercially available because it was found to improve the properties of
hardened concrete. Inhibitors with sodium or potassium bases were found to be detrimental to concrete
strength and to cause alkali-aggregate reaction. In 1983, the Federal Highway Administration concluded
that using calcium nitrate as an admixture could provide more than an order of magnitude reduction in the
corrosion rate. More recent tests have found that the use of calcium nitrite as an admixture will provide
corrosion protection for diffused chlorides up to a level of 16 lb/yd3 (Berke 1991). This is substantially
higher than the chloride threshold of 1 lb/yd3.
The calcium nitrite delays the onset of corrosion initiation and controls the rate of corrosion by stabilizing
the passivating layer of iron-oxide film. It chemically reacts with the passivating layer of the embedded
steel so that it remains intact when it comes into contact with chlorides (Perkins 1997). Calcium nitrite is
an anodic inhibitor. It prevents corrosion activity by suppressing the anodic reaction.
The use of a calcium nitrite admixture is limited to new construction. For rehabilitation, corrosion
inhibitors are available in the form of a surface applied liquid. These products migrate through the
concrete surface in order to provide a protective film on the reinforcing steel. Some surface applied
corrosion inhibitors are calcium nitrite based. Since nitrites pose an environmental threat, many of the
newer surface applied corrosion inhibitors use organic and inorganic materials instead. Most are based on
amine salts. Some are water based and some are alcohol based. Noncalcium nitrite based inhibitors are
usually mixed inhibitors in that they act as both anodic and cathodic inhibitors. In addition to providing a
protective film on the anode, they suppress the cathodic reaction by forming a barrier at the cathodic site
that prevents oxygen from reaching the steel (Shaw 1997).
Laboratory testing of surface applied corrosion inhibitors has concluded that they can reduce corrosion.
However, most of these tests have been sponsored by the product manufacturers and have also been
performed on new concrete. Very little is known about the performance of corrosion inhibitors for
repaired construction. The effect the corrosion inhibitor has on the reinforcing outside of the repair area is
unknown. In addition, little is known about how well it protects steel that is actively corroding, or under
conditions that cause macrocells to form (Krauss, Gu, and Vaysburd 1999). Since surface applied
corrosion inhibitors have only been available since the 1990s, results from laboratory and field testing are
limited. There have been no systematic studies reported on the effectiveness of surface applied corrosion
inhibitors on existing structures or chloride contaminated concrete.

69

8.3

DESCRIPTION OF TEST SPECIMENS WITH CORROSION INHIBITORS

In TxDOT Project 1774, an additional laboratory study was designed to investigate the effectiveness of
three commercially available surface applied corrosion inhibitors. The preliminary findings from the
laboratory study discussed in Chapter 6 indicated that the use of FRP composite wrap alone on chloride
contaminated concrete did not prevent corrosion activity. The objective of the additional test program is
to evaluate the effectiveness of FRP composite wrap in the prevention of corrosion activity when used in
conjunction with a surface applied corrosion inhibitor. For consistency with the initial study, similar
specimens were fabricated. The properties for each specimen are listed in Table 8.1.

Table 8.1 Specimen Parameters for Corrosion Inhibitor Study

Specimen #

Concrete
Condition

Corrosion
Inhibitor

Wrap Length (in)

Cracked

Surtreat

24

Cracked

Surtreat

36

Uncracked

Surtreat

24

Uncracked

Surtreat

36

Cracked

Cortec

24

Cracked

Cortec

36

Uncracked

Cortec

24

Uncracked

Cortec

36

Cracked

Sika

24

10

Cracked

Sika

36

11

Uncracked

Sika

24

12

Uncracked

Sika

36

13

Cracked

None

24

14

Cracked

None

36

15

Uncracked

None

24

16

Uncracked

None

36

17

Cracked

None

None

Probe
Installation

VETEK

VETEK

VETEK

VETEK

VETEK

The specimen geometry, cast-in-chloride concrete mix, and cracking conditions for the specimens in this
study are the same as the ones described in Chapter 5. Only cylinder (column) specimens were used.
One FRP composite system TYFO S Fibrwrap was used because it is commercially available and is
being used on TxDOT projects. The specimens were wrapped either to the waterline with a 24 in. partial
wrap, or a 36 in. full length wrap.
Three commercially available surface applied corrosion inhibitors were selected. The products as
described in the manufacturers datasheets are summarized below:

Total Performance System (TPS), manufactured by Surtreat International. TPS uses a water
soluble chemical formulation that controls the pH level in concrete and ties up chlorides and
drives salts to the surface of the concrete.
70

MCI 2020, manufactured by Cortec Corp. MCI 2020 is a water-based blend of surfactants and
amine salts that migrates to the steel reinforcing. It forms a monomolecular protective layer on
the reinforcing steel.

FerroGard-903, manufactured by Sika Corp. Ferrogard 903 is a modified amino alcohol

inhibitor that migrates to the steel reinforcing in order to form a thin protective coating on the
steel surface. The film inhibits corrosion and displaces chlorides.

In constructing the previous cylinder specimens, poor consolidation at the time of casting resulted in
severe honeycombing. For the new specimens, procedures were changed to improve consolidation. The
formwork consisted of 36 in. high cardboard sonoforms that were firmly anchored to prevent them from
floating up when the concrete was cast. Twenty-six specimens were cast. The temperature was 65F
during the time of placement. The concrete slump was 6 -in. The concrete was placed into the form
from an overhead chute in one lift. An immersion-type vibrator was placed in the form prior to the
concrete placement and slowly pulled out as the concrete was added in order to assure good consolidation
throughout the specimen. After the concrete was placed, the top surface was struck off and troweled. The
specimens were covered with heavy damp clothes and a plastic sheet to cure for 28 days. All but one of
the specimens were well consolidated and had smooth finishes. The average 28 day concrete cylinder
strength was 3520 psi.
Half of the specimens were cracked as desrcibed in Chapter 5. The specimens with the most uniform
cracking patterns were selected for the experiement.
The application of the corrosion inhibitors and the FRP wrap was done by Delta Structural Technology,
Inc. The surface of the specimens was cleaned off with an air hose prior to application of the surface
applied corrosion inhibitors. The inhibitors were then applied with a low pressure sprayer according to the
manufacturers instructions. A separate sprayer was used for each type of inhibitor. All of the corrosion
inhibitors were applied at a rate of 100 ft2/gal. The Surtreat TPS was applied using three coats. The
specimens were subjected to a light rinse after the second and third application. The Cortec MCI 2020
was applied using two coats and rinsed after each coat. The Sika Ferrogard 903 was applied using three
coats and rinsed after each coat. The specimens that were not being treated were covered with a plastic
drop cloth to prevent contamination. Figure 8.1 shows the application of the corrosion inhibitor.

Figure 8.1 Application of a surface applied corrosion inhibitor to the specimens

71

Before the wrap was applied, the surface of the specimens were prepared by applying an epoxy filler
thickened with Cab-O-Sil, to create a smooth even surface. The glass fabric was saturated with the
epoxy resin using a paint roller. The wrap was then applied to the specimens using the hand layup
method. Three layers were applied with a 6 in. overlap. After each layer was applied, the surface was
smoothed out by hand in order to remove air pockets. The specimens were cured for 24 hours and then
inspected for voids under the wrap. Any voids detected were injected with epoxy. The specimens were
painted with Sherwin Williams Hi Bald Aliphatic Polyurethane paint. Afterwards, they were left to cure
an additional six days, bringing the total curing time to seven days. Figures 8.2 through 8.4 show the
application of the FRP composite wrap.

Figure 8.2 Filling voids with a thickened epoxy filler

Figure 8.3 Saturating glass fabric with epoxy resin

72

Figure 8.4 Applying FRP wrap to the concrete specimen

8.4

MONITORING AND EXPOSURE

The specimens will be monitored using half-cell potentials. The specimens were prepared for taking
readings in the wrapped sections in the same manner as described in Chapter 5. In addition to taking halfcell potential measurements at the concrete surface, corrosion monitoring probes were embedded half-cell
probes were installed in five of the specimens. By embedding the probes, the potential measurements are
taken at the steel rather than at the concrete surface. This reduces error related to properties of the
concrete such as moisture content, thickness of the concrete cover, and resistivity of the concrete. It also
offers the added benefit that measurements may be taken at any time without having to unseal a portion of
the FRP composite.
The VETEK System manufactured by Corrosion Monitoring Systems was chosen for the embedded halfcell probes. The system consists of two separate electrodes. Both electrodes are wrapped in a permeable,
nonconducting PVC covering. The probes monitor the steel for a distance of approximately 10 cm from
the probe location. A separate connection to the steel must be made for each reinforcing bar being
monitored. The V2000 Monitoring Electrode consists of a solid silver/silver chloride wire electrode. The
V1500 Monitoring Electrode is a pure gold wire electrode. Both electrodes monitor corrosion activity,
and when used together they provide information on the chloride concentration in the concrete. The
chloride concentration is found by taking the difference in potential readings between the gold electrode
and the silver electrode and plotting it on the graph in Figure 8.5. The manufacturers guide for
interpreting the results for each probe is summarized in Table 8.2 (CMS 2000).

Table 8.2 Interpretation of Readings for the VETEK System

V2000 (silver)
Ecorr (mV)

V1500 (gold)
Ecorr (mV)

Corrosion Risk

0 to 300

100 to -150

No active corrosion in the vicinity of the probe

300 to 400

-150 to 200

Damage of the passive layer has begun

> 400

< -200

Active corrosion is taking place

73

Gold vs Silver/Silver Chloride Electrodes

0

Potential, mV

-50
-100
-150
-200
-250

Potential, mV

-300
0.001

0.01

0.1

10

Chloride Concentration, mol/liter

Figure 8.5 Chloride determination of VETEK System

The silver/silver chloride probe may be converted to an equivalent copper/copper sulfate value by
subtracting the reading from 94 mV (Broomfield 1997).
The probes were installed by drilling a 1 in. hole into the concrete up to the steel reinforcing. The probes
were then bent into a circle and placed in the hole. Care was taken to prevent the wires in the two
electrodes from coming into direct contact, which could result in an electrical short. After the probes were
placed, the hole was then filled with a sand/cement concrete mix as specified by the manufacturer. The
location of the installation was 4 in. above the waterline. A connection was made to the steel reinforcing
bar being monitored by using the steel connector that was included with the system. The installation of
the VETEK system is shown in Figures 8.6 and 8.7.

Figure 8.6 Installation of the embedded electrodes

74

Figure 8.7 Wires from the embedded probes and the connection to the steel
Unlike traditional half-cell measurements, the electrical connection to the steel reinforcing is made to the
negative terminal of the voltmeter. The reference electrode is connected to the positive terminal of the
voltmeter. This results in readings that are positive rather than negative.
Before the specimens were placed in the exposure tank, half-cell potential measurements were taken for
each specimen. The results are listed in Tables 8.3 through 8.5.
There is some general agreement between the half-cell potential readings taken at the concrete surface
and the half-cell potential readings taken from the embedded probe. The Ecorr values for the embedded
probe readings are typically more negative than Ecorr for the standard half-cell potential when both values
were converted to the copper/copper sulfate equivalents. There is a distinct possibility that a macrocell
may have formed at the location of the embedded probe due to the difference in chlorides between the
concrete and the concrete grout that was used to fill in the hole after the probes were placed. Also the
half-cell measurements that are taken at the concrete surface are taken on a different reinforcing bar than
the one that is being monitored with the embedded probes. Therefore different levels of corrosion activity
may be taking place on each bar at the location of 4 in. above the waterline.
The specimens treated with the corrosion inhibitors appear to have lower chloride levels than the
untreated specimens. The measurements indicated that the specimens treated with the Cortec and Sika
corrosion inhibitors had substantially lower chloride levels. All of the specimens were fabricated with the
same chloride contaminated concrete and the same non-chloride concrete grout, so the chloride levels
were expected to be the same. The half-cell potential values for the specimens that were not treated with
the corrosion inhibitors were generally less negative than the specimens that were treated.
The new specimens, along with the remaining specimens from the initial study, were placed in a new
exposure tank. The exposure tank that had been used for the study described in Chapters 5 and 6 had
started experiencing problems with leaking. The new exposure tank was built out of concrete and lined
with epoxy to prevent leaking and corrosion of the reinforcing in the tank. Figure 8.8 is a photograph of
the exposure environment.

75

Table 8.3 Half-cell Potential Readings

Specimen

Ecorr (mV)

Interpretation

-212

Intermediate Risk

-225

Intermediate Risk

-211

Intermediate Risk

-295

Intermediate Risk

-298

Intermediate Risk

-243

Intermediate Risk

-186

Low Risk

-306

Intermediate Risk

-238

Intermediate Risk

10

-200

Low Risk

11

-229

Intermediate Risk

12

-279

Intermediate Risk

13

-179

Low Risk

14

-244

Intermediate Risk

15

-184

Low Risk

16

-262

Intermediate Risk

17

-100

Low Risk

Table 8.4 Embedded Probe Readings

Specimen

Gold
probe
(mV)

Silver
probe
(mV)

VETEK
interpretation of
Corrosion
Activity

Cu/CuSO4
Equivalent
(mV)

ASTM
Interpretation of
Corrosion Activity

183

190

No corrosion

-284

Uncertain

435

Corrosion

-529

Corrosion

10

-6

186

No corrosion

-280

Uncertain

14

No corrosion

-97

Low probability

17

No corrosion

-94

Low probability

76

Table 8.5 Chloride Measurements from Embedded Probes

Specimen

Difference between
probe reading

Chloride Level
(mols/L)

Interpretation

-7

32

High levels

-427

Low levels

10

-192

0.02

Low levels

14

53

High levels

17

42

High levels

Figure 8.8 Specimens in the exposure tank.

The specimens will continue to be exposed to wet/dry cycles consisting of the one week wet followed by
two weeks air drying. Half-cell potential readings are planned after every cycle for the embedded probes
and after every four cycles for the surface half-cell potential readings.

77

78

CHAPTER 9: SUMMARY AND CONCLUSIONS

9.1

SUMMARY

Corrosion of steel reinforcing in concrete is one of the greatest threats to the durability of reinforced
concrete. It is the single most expensive corrosion related problem, and affects the integrity of thousands
of reinforced concrete structures. Bridges located in aggressive environments, such as exposure to
seawater or deicing salts are highly susceptible to corrosion activity.
Fiber reinforced plastic (FRP) composite wraps have recently been suggested to be used in repair of
damage to concrete that resulted from corrosion activity. FRP composite wraps have been proven to
strengthen concrete members and improve ductility. Over the last decade, they have performed well in
seismic retrofit applications.
Little is known about the long-term effectiveness of FRP composite wrap in prevention of corrosion in
reinforced concrete. The FRP may act as a barrier to protect the concrete from moisture and chlorides,
which can lead to corrosion, or it may trap already existing moisture and chlorides in the concrete,
allowing for the corrosion process to continue undetected.
In this study, laboratory specimens that represent conditions present in chloride contaminated concrete
were exposed to an accelerated aggressive environment. The specimens were monitored and evaluated for
corrosion activity. In addition, bridge overpasses that have been wrapped with FRP after experiencing
corrosion related damage were monitored for corrosion activity.

9.2

REVIEW OF FINDINGS

The following observations were noted in the laboratory and the field.
9.2.1 Corrosion Activity
The monitoring methods used in this study: half-cell potential, linear polarization, and embedded probes
indicated that corrosion activity was likely in all of the specimens in the study. Corrosion activity
generally corresponded with areas that had the most severe exposure to wetting and drying; typically in
the damaged portions of the concrete, especially at crack locations.
9.2.2 Role of FRP as a Barrier
The FRP composite wrap did provide a barrier for chlorides and moisture for the downstream ends of the
rectangular block specimens and the upper halves of the cylinder specimens. Very little corrosion activity
was found in those areas for the concrete specimens without the cast-in-chlorides. Corrosion activity was
found in those locations for the chloride contaminated specimens. In the case of the cylinder specimens,
increased chlorides were found in the splash zone. This included a cylinder specimen that was fully
wrapped and did not contain cast-in-chlorides. In addition, trapped moisture was found underneath the
wrap in four out of five of the fully wrapped specimens when the wrap was removed. The moisture was
found near the edge of the wrap, since the specimens were not fully encapsulated Accelerated corrosion
was also found near the locations of the trapped moisture. There was some section loss of a steel stirrup in
one of the specimens.
9.2.3 Effect of Repair Materials
Corrosion activity was further accelerated at locations of patches of repair material in the chloride
contaminated concrete. Corrosion was also found near the patch location in the downstream end of a fully
wrapped rectangular block that did not contain cast-in-chlorides. The activity was minor compared to
corrosion activity taking place near the patch locations of the chloride contaminated concrete.
79

9.2.4 Field Observations

In the field study, high chloride levels were found in portions of concrete that were not showing signs of
damage due to corrosion at the time repairs were made. The linear polarization testing indicates that low
to moderate corrosion activity is most likely taking place in the structures after the members were
wrapped. It was observed that some surface areas on the tops of the bridge bents were left exposed near
the edge of the wrap around the bearing pads. Corrosion rates were higher in such locations compared
with areas where the surface was completely sealed. Also, one bridge bent had visible signs of corrosion
activity taking place. Rust stains were found on the underside of the bent at the seams in the FRP wrap.

9.3

PRELIMINARY CONCLUSIONS AND RECOMMENDATIONS

The following conclusions and recommendations were drawn from this study.
9.3.1 Effectiveness of FRP in Corrosion Prevention
Wrapping of chloride contaminated concrete does not appear to prevent corrosion. There is a serious risk
that corrosion may continue in areas where moisture may infiltrate, and also in areas with dissimilar
repair materials. Macrocells may develop between existing chloride contaminated concrete and the repair
material even in areas where moisture infiltration is unlikely. However, very little corrosion activity was
found in the wrapped portions of chloride free, undamaged concrete. This is most likely due to the ability
of the FRP composite wrap to act as a barrier to chlorides and moisture.
9.3.2 FRP as a Barrier
In the laboratory study, the FRP composite wrap appeared to provide a barrier to chlorides and moisture.
However, moisture was able to enter the wrapped portions of concrete through adjacent exposed concrete
surface. Increased chlorides and trapped moisture was found in the concrete near the edge of the wrap. It
does not appear to be feasible to totally wrap bridge members in the field.
9.3.3 Corrosion Monitoring
The half-cell potential, linear polarization, and embedded probes all proved to be viable options for
evaluating corrosion activity. Linear polarization is the only method that establishes the degree of
corrosion activity taking place. The half-cell potential is limited to determining whether corrosion activity
is taking place. In a laboratory environment where corrosion activity is expected, half-cell potential
readings provide little information after the onset of corrosion.
Because FRP composite wraps obscure the concrete surface, all corrosion monitoring is limited to
predetermined locations where access to the concrete is provided. As a result, corrosion activity may
appear to be more severe than it actually is, or corrosion activity may take place undetected. The
corrosion activity at the location being monitored may not necessary reflect conditions a short distance
away.
9.3.4 Recommendations for Field Applications
It is recommended that all chloride contaminated concrete should be removed before FRP composite wrap
is applied when repairing concrete damage due to corrosion activity. It is also recommended that care
should be taken to seal exposed concrete near the edge of the wrap on structures that are exposed to
deicing salts. The less exposed concrete, the lower the amount of moisture and chlorides that are likely to
enter bridge members. In marine environments, preventing moisture and chlorides from entering bridge
columns is an area of concern. It may be difficult to prevent capillary action from taking place. In
addition, it is recommended that embedded probes be installed in areas that may be susceptible to
corrosion activity in order to monitor the structure.

80

9.4

RECOMMENDATIONS FOR CONTINUING RESEARCH

Further research is needed into the long-term effects of corrosion activity of reinforcing steel in FRP
composite wrapped concrete. While corrosion activity has been found in the specimens examined, the
extent of corrosion activity over a longer period of time is unknown.
In addition to long-term studies, more research is needed to assess the effectiveness of FRP composite
wrap in preventing or reducing corrosion activity in concrete that has been repaired properly. Adequate
preparation includes the removal of all chloride contaminated concrete and sealing of cracks. The use of
surface applied corrosion inhibitors needs further exploration, especially where chloride contaminated
concrete is involved.

81

82

Appendix A: Details of Test Specimens

Table A.1 Parameters for Chloride Contaminated Cylinders

FRP Wrap
Specimen*

Fabric type

CC1
CC2
CC3
CC4

Fibrwrap
Generic
Fibrwrap
Fibrwrap

CC5
CC6
CC7
CC8
CC9
CC10

Generic
Generic
Fibrwrap
Fibrwrap
Fibrwrap
None

CC11
CC12
CC13
CC14
CC15
CC16
CC17

None
Generic
Generic
Generic
Generic
None
None

CC18
CC19
CC20
CC21

None
Generic
Generic
None

Resin
system

Initial
concrete
condition

Length,
in
24
30
24
24

TYFO S
TYFO S
TYFO S
TYFO S

Cracked
Wet

36
36
24
36
24

Epoxy
Vinyl Ester
TYFO S
TYFO S
TYFO S

Cracked
Cracked
Cracked
Cracked

Concrete
repair
material+

Corrosion
Inhibitor
Ferrogard

LMC
EG
LMC
EG
EG

Ferrogard

LMC
Ferrogard

30
24
24
24

Epoxy
Epoxy
Epoxy
Epoxy

Wet
Cracked
LMC
Cracked

Ferrogard
Ferrogard

EG
LMC
Cracked
24
24

Vinyl Ester
Vinyl Ester

LMC
Cracked

83

Ferrogard
Ferrogard

Table A.2 Parameters for Non-Chloride Cylinders

FRP Wrap
Specimen*

Resin
system

Fabric type

Length,
in

CNC1

Generic

27

Epoxy

CNC2

Generic

36

Epoxy

CNC3
CNC4
CNC5

Generic
Fibrwrap
Fibrwrap

24
24
36

Epoxy
TYFO S
TYFO S

CNC6

Generic

24

Vinyl Ester

CNC7
CNC8
CNC9
CNC10
CNC11

None
None
Generic
Fibrwrap
None

CNC12
CNC13
CNC14
CNC15
CNC16
CNC17

None
Generic
Generic
None
Fibrwrap
Fibrwrap

CNC18
CNC19
CNC20

Generic
Generic
None

Initial
concrete
condition
Cracked,
Wet
Cracked

Concrete
repair
material+
EG

Ferrogard
Cracked
Cracked,
Wet

EG
Ferrogard
Ferrogard

Cracked
24
24

Vinyl Ester
TYFO S

Corrosion
Inhibitor

LMC
Cracked
LMC
EG

24
36

Epoxy
Epoxy

24
24

TYFO S
TYFO S

LMC
EG

24
24

Epoxy
Epoxy

LMC

84

Cracked
Cracked
Cracked

Ferrogard
Ferrogard

Ferrogard

Table A.3 Parameters for Chloride Contaminated Rectangular Blocks

FRP Wrap
Specimen*

Fabric type

Length,
in
27

Resin
system

RC1

Generic

RC2
RC3
RC4
RC5
RC6

Generic
Fibrwrap
None
Fibrwrap
Fibrwrap

31
24

Vinyl Ester
TYFO S

27
33

TYFO S
Epoxy

RC7
RC8
RC9

Generic
None
Fibrwrap

30

Epoxy

Initial
concrete
condition

Epoxy

Concrete
repair
material+

Corrosion
Inhibitor

LMC

Ferrogard

Cracked
Cracked
Cracked
LMC
LMC
Cracked
LMC

24

Epoxy

Cracked

Ferrogard

Table A.4 Parameters for Non-Chloride Rectangular Blocks

FRP Wrap
Specimen*

Fabric type

RNC1
RNC2
RNC3
RNC4

Fibrwrap
None
Generic
Generic

RNC5
RNC6
RNC7
RNC8

Fibrwrap
Fibrwrap
None
Generic

Resin
system

Length,
in
24

TYFO S

27
36

Epoxy
Vinyl Ester

30
3030

TYFO S
Epoxy

24

Epoxy

Initial
concrete
condition

Concrete
repair
material+

Corrosion
Inhibitor

Ferrogard
Cracked
Cracked
Cracked
Cracked
Cracked

LMC
LMC

* The first letter represents the specimen geometry (C for cylinder, R for rectangular block). The
following letters denote which concrete mix was used (C for cast-in-chlorides, NC for no
chlorides).
+

The notation EG is for the epoxy grout, and the notation LMC is for the latex-modified
concrete.

85

86

Appendix B: Half-Cell Potential Readings 2/99-4/00

Table B.1
Date

CC18

CNC8

02/26/99

-18

-66

-24

05/13/99

-221

-326

-377

06/11/99

-310

-333

-307

07/02/99

-340

-361

-380

07/28/99

-251

-347

-387

08/25/99

-411

-417

-447

09/15/99

-443

-437

-423

-501

-513

11/03/99

-544

-565

-572

11/26/99

-546

-584

-574

12/15/99

-579

-586

-574

01/06/00

-683

-521

-606

-528

-584

02/29/00

-610

-659

-607

03/25/00

-616

-674

-615

-590

-667

10/06/99

01/28/00

04/18/00

CC7

Half-cell Potential Readings for Group A

(mV vs. Cu/CuSO4)

-547

-588

-608

CNC13 CNC14 CNC19

-574

-597

-611

87

-425

-507

-535

RC4

RC7

-471

-507

-556

-566

-609

-386

-397

-411

Table B.2
Date

CC3

Half-cell Potential Readings for Group B

(mV vs. Cu/CuSO4)
CC5

CC6

CNC10

RNC6

RNC7

02/26/99

-32

05/13/99

-392

06/11/99

-398

07/02/99

-487

07/28/99

-486

08/25/99

-476

09/15/99

-499

10/06/99

-288

-589

-583

-449

11/03/99

-276

-589

11/26/99

-589

12/15/99

-585

01/06/00

-606

01/28/00

-355

-606

-580

-524

-375

-555

02/29/00

-620

03/25/00

-414

04/14/00

-670

05/08/00

-640

06/07/00

-409

-638

-658

-560

-488

-645

07/03/00

-588

07/26/00

-621

08/17/00

-640

10/15/00

-255

-513

-510

-394

-316

-547

02/05/01

-212

-460

-459

-356

-228

-462

88

Appendix C: PHOTOGRAPHS OF STEEL REINFORCING

Group A

Figure C.1 Specimen CC7

Figure C.2 Specimen CC18

89

Figure C.3 Specimen CNC8

Figure C.4 Specimen CNC13

90

Figure C.5 Specimen CNC14

Figure C.6 Specimen CNC19

91

Figure C.7 Specimen RC4

Figure C.8 Specimen RC7

92

Group B

Figure C.9 Specimen CC3

Figure C.10 Specimen CC5

93

Figure C.11 Specimen CC6

Figure C.12 Specimen CNC10

94

Figure C.13 Specimen RNC6

Figure C.14 Specimen RNC7

95

96

References
Abu-Tair, A.L., E. Burley, and S.R. Rigden. (1995). Evaluating the Pull-Off Bond Strength of
Three Repair Materials Subjected to Different Loading and Exposure Conditions,
Proceedings of the Fifth International Conference, Structural Faults and Repairs, Volume 2,
Engineering Technics Press, Edinburgh, UK, pp. 355-364.
Allen, R.T.L., S.C. Edwards, and J.D.N. Shaw, ed. (1993). The Repair of Concrete Structures,
Second Edition, Blackie Academic & Professional, Glasgow, UK.
American Concrete Institute. (1995). Building Code Requirements for Structural Concrete (ACI
318-95) and Commentary (ACI 318R-95), American Concrete Institute, Farmington Hills,
MI.
American Concrete Institute. (1992). Control of Cracking in Concrete Structures, 224R-92,
American Concrete Institute, Detroit, MI.
American Society of Civil Engineers. (2001). Report Card for Americas Infrastructure:
Bridges, American Society of Civil Engineers.
ASTM. (1989). Standard Test Method for Half-Cell Potential of Uncoated Reinforcing Steel in
Concrete, Annual Book of ASTM Standards, Designation: C876-87, ASTM, West
Conshohocken, PA, pp. 428-432.
Babaei, Khossrow and Neil M. Hawkins. (1991). Evaluation of Bridge Deck Protective
Strategies, Concrete Durability: Corrosion Protection, ACI Compilation 25, American
Concrete Institute, Detroit, MI, pp. 18-28.
Ballinger, Craig A. (1991). Development of Composites for Civil Engineering, Advanced
Composites Materials in Civil Engineering Structures, American Society of Civil Engineers,
NY.
Bassett, Susan. (1998). Composites for Infrastructure: A guide for Civil Engineers, Ray
Publishing, Wheat Ridge, CO.
Berke, Neal S. (1991). Corrosion Inhibitors in Concrete, Concrete Durability: Corrosion
Protection, ACI Compilation 25, American Concrete Institute, Detroit, MI, pp. 14-17
Broomfield, John P. (1997). Corrosion of Steel in Concrete, E.& F.N. Spon, London, UK.
Concorr. (1998). Corrosion Rate Probe and Connection Detail, Concorr, Inc., Ashburn, VA.
Corrosion Monitoring Systems. (2000). VETEK Product Information.
Cortec Corporation. (2000). MCI-2020 Product Data Sheet.
CRC Handbook on Nondestructive Testing of Concrete. (1991). CRC Press, Boca Raton, FL.

97

Delta Structural Technology, Inc. (2000). Composite Fiberwrap, Delta Structural Technology,
Inc., Humble, TX.
Emmons, Peter H., Alexander M. Vaysburd, and Jay Thomas. (1998). Strengthening Concrete
Structures, Part II, Concrete International, April, pp. 56-60.
Erki, M.A. and S.H. Rizkalla. (1993). FRP Reinforcement for Concrete Structures, Concrete
International, June, pp. 48-53.
Feld, Jacob and Kenneth L. Carper. (1997). Construction Failure, Second Edition, John Wiley &
Sons, Inc., NY.
Ferguson Structural Engineering Laboratory. (1999). Effect of Wrapping Chloride
Contaminated Structural Concrete With Layers of Glass Fiber/Composites and Resin,
Center for Transportation Research, Austin, TX.
Fuentes, Laura Alexandra. (1999). Implementation of Composite Wrapping Systems on
Reinforced Concrete Structures Exposed to a Corrosive Laboratory Environment, Thesis,
Civil Engineering Department, University of Texas at Austin.
Fyfe, Edward R., (1995). Testing and Field Performance of High Strength Fiber Wrapping
System, Restructuring: America and Beyond, Proceedings of Structures Congress XIII,
Volume 1, pp. 603-606.
Gaynor, R. (1987). Understanding Chloride Percentages, Corrosion, Concrete, and Chlorides.
Steel Corrosion in Concrete: Causes and Restraints, (SP102-11), American Concrete
Institute, Detroit, MI, pp. 161-165.
Gu, Ping, et al. (1997). Electrochemical Incompatibility of Patches in Reinforced Concrete,
Concrete International, August, pp. 68-72.
Hansson, Carolyn M. and Birgit Sorensen. (1990). The Threshold Concentration of Chloride in
Concrete for the Initiation of Reinforcement Corrosion, Corrosion Rates of Steel in
Concrete, ASTM STP 1065, Philadelphia, PA, pp. 3-16.
James Instruments, Inc. Test CL-500 To Determine the Total Content of Chlorides in Concrete,
James Instruments, Inc., Chicago, IL.
John, D.G., et al. (1995). Corrosion Management of Reinforced & Pre-stressed Concrete
Bridges Using Permanently Embedded Sensors, Proceedings of the Fifth International
Conference, Structural Faults and Repairs, Volume 1, Engineering Technics Press,
Edinburgh, UK, pp. 123-128.
Jones, Denny A. (1996). Principles and Prevention of Corrosion, Second Edition, Prentice Hall,
Upper Saddle River, NJ.
Karbhari, Vistasp M., Frieder Seible, and Gilbert A. Hegemier. (1996). On the Use of Fiber
Reinforced Composites for Infrastructure Renewal- A Systems Approach, Materials for the
New Millennium, American Society of Civil Engineers, NY, pp. 1091-1100.

98

Kay, Ted. (1992). Assessment and Renovation of Concrete Structures, Longman Scientific &
Technical, NY.
Krauss, Paul, Gu, Gordon Ping, and Alexander Vaysburd. (1999). Perspectives on Corrosion
Inhibitors, Repair Perspectives, Publication #R970154, The Aberdeen Group.
Labossiere, P., Neale, K.W., and Martel, S. (1997). Strengthening Existing Structures With
Composite Materials: Practical Applications in Quebec, Recent Advances in Bridge
Engineering, July, pp. 89-97.
Leeming, M.B. and V. Peshkam. (1995). Robust Solution to Strengthening Bridges,
Proceedings of the Fifth International Conference, Structural Faults and Repairs, Volume 1,
Engineering Technics Press, Edinburgh, UK, pp. 161-164.
Long, A.E. and A. McC. Murray. (1984). The Pull-Off Partially Destructive Test for
Concrete, In Situ/Nondestructive Testing of Concrete, (SP 82-17), American Concrete
Institute, Detroit, MI, pp. 327-350.
Mailvaganam, Noel P., ed. (1992). Repair and Protection of Concrete Structures, Institute for
Research in Construction, National Research Council of Canada, CRC Press, Inc., Ottawa,
Ontario.
Mallick, P. K. ed. (1997). Composites Engineering Handbook, University of MichiganDearborn, Dearborn, MI.
Mirmiran, Amir and Mohsen Shahawy. (1997). Behavior of Concrete Columns Confined by
Fiber Composites, Journal of Structural Engineering, May, pp. 583-590.
Nanni, Antonio. (1995). Concrete Repair With Externally Bonded FRP Reinforcement,
Concrete International, June, pp. 22-26.
Neale, Kenneth W. and Pierre Labossiere. (1998). Fiber Composite Sheets in Cold Climate
Rehab, Concrete International, June, pp. 22-24.
Ohta, Toshitaka, et al. (1992). Deterioration in a Rehabilitated Prestressed Concrete Bridge,
ACI Materials Journal, July-August, pp. 328-336.
Ohtsu, Masayasa, Toshifumi Yamamoto, and Kimitoshi Matsuyama. (1997). Quantitative NDE
Estimation for Rebar Corrosion in Concrete, Proceedings of the Seventh International
Conference, Structural Faults and Repairs, Volume 2, Engineering Technics Press,
Edinburgh, UK, pp. 265-273.
Perkins, P.H. (1997). Repair, Protection and Waterproofing of Concrete Structures, E.& F.N.
Spon, London, UK.
Saadatmanesh, H., M.R. Ehsani, and M.W. Li. (1994). Strength and Ductility of Concrete
Columns Externally Reinforced with Fiber Composite Straps, ACI Structural Journal, JulyAugust, pp. 434-447.

99

Scannell, William T., Ali A. Sohanghpurwala, and Moavin Islam. (1996). Assessment of
Physical Condition of Concrete Bridge Components, FHWA-SHRP Showcase Participants
Workbook, Concorr, Inc., Ashburn, VA .
Shahawy, Mohsen and Thomas E. Beitelman. (1995). Structural Applications of Advanced
Composite Materials in Bridge Construction and Repair, Restructuring: America and
Beyond, Proceedings of Structures Congress XIII, Volume 1, pp. 607-621.
Shaw, Mark. (1997). Migrating Corrosion Inhibitors fro Reinforced Concrete Protection,
Proceedings of the Seventh International Conference, Structural Faults and Repairs, Volume
2, Engineering Technics Press, Edinburgh, UK, pp. 317-324.
Sherwin Williams (2001). Macropoxy 920 Pre-Prime Data Sheet.
Sheikh, S. et al. (1999) Repair of Delaminated Circular Pier Columns by ACM., Ontario Joint
Transportation Research Report, MTO Reference Number 31902.
Sika Corporation. (2000). Sika FerroGard 903 Product Data Sheet.
Sika Corporation. (1996). Construction Products Catalog.
Sohanghpurwala, Ali and William T. Scannell. (1994). Repair and Protection of Concrete
Exposed to Seawater, Concrete Repair Bulletin, July-August, pp. 8-11.
Soudki, Khaled A. and Mark F. Green. (1997). Freeze-Thaw Response of CFRP Wrapped
Concrete, Concrete Interational, August, pp. 64-67.
Surtreat International. (2000). Total Performance System Technical Data.
Tang, Benjamin. (1997). Fiber Reinforced Polymer Composites Applications in USA, First
Korea/U.S.A. Road Workshop Proceedings.
Teng, Ming-Hung, Elisa D. Sotelino, and Wang Chen. (2000). Monitoring of Long-Term
Performance of Highway Bridge Columns Retrofitted by Advanced Composite Jackets in
Indiana, Joint Transportation Research Program, West Lafayette, Indiana.
Thomas, Jay, et al. (1996). Externally Bonded Carbon Fiber For Strengthening Concrete,
Materials for the New Millennium, American Society of Civil Engineers, NY, pp. 924-931.
Toutanji, H. and P. Balaguru. (1998). Durability Characteristics of Concrete Columns Wrapped
with FRP Tow Sheets, Journal of Materials in Civil Engineering, February, pp. 52-57.
Verhulst, Stewart Mason. (1999). Evaluation and Performance Monitoring of Corrosion
Protection Provided by Fiber-Reinforcing Wrapping, Thesis, Civil Engineering Department,
University of Texas at Austin.
Watson, Ronald J. (2000). Wrapping it Up, ASTM Standardization News, February, pp. 22-25.
Whitmore, David, Sean Abbott, and Emmanuel Velivasakis. (1999). Battling Concrete
Corrosion, Civil Engineering, January, pp. 46-48.
100

Yeomans, S. R. (1991). Comparative Studies of Galvanized and Epoxy Coated Steel

Reinforcement in Concrete, Durability of Concrete, Second International Conference,
Volume I, (SP 126-19), American Concrete Institute, Detroit, MI, pp. 355-370.

101